What is the Percentage of Homeschoolers in the US?

District budget staff in a growing county now track homeschool notices like kindergarten registration. A one-point swing can erase a classroom of funding. A parent at the kitchen table faces a different problem. One official file says about 3.4% of students are homeschooled. Another says about 6%. The notice of intent depends on which cut describes her kid. Those two jobs keep hitting the same messy spreadsheet. What follows separates the 2024-2025 share from the headcount. Then it moves to older federal surveys, state ranks, family mix, money, test scores, and the 2025-2026 statute fights.

What Percentage of US K-12 Students Are Homeschooled?

• The working national figure for 2024-2025 is 6.262% of school-age children. It blends Census Household Pulse results and state registration files after undercount adjustments.
• U.S. Census Bureau Household Pulse data collected August 20-September 16, 2024 put the raw homeschool share at 6.530% of school-age children.
• A 23-state registration mean compiled by the Johns Hopkins Homeschool Hub came in much lower: 4.18%. That mean is weighted by each state’s homeschool count and school-age population.
• Raise that state-file mean by a 20% underrepresentation adjustment and it becomes 5.016%.
• Apply a 15% underrepresentation adjustment to the Pulse share and it becomes 7.509%. So the 6.262% figure is the mean of those two adjusted cuts. It is not a third independent census.
• A 2026 sector-share compilation mixed current state files with lagged ones. It put homeschooling at 4.8% of K-12 students.
• Averaging Pulse weeks across 2023-2024 produced 5.92% homeschooled, versus 5.82% in 2022-2023.
• Weekly Pulse readings in 2023-2024 ran from 5.20% to 6.91%. Because of that spread, a single-week screenshot can look like a boom or a stall.
• Treat any single percentage as a cut, not the cut. Student-share surveys, household-share pulses, and registrar files are not measuring the same kids.

How Many US Students Are Homeschooled?

• The matching 2024-2025 headcount is 3.408 million K-12 homeschool students.
• A ±10% band around that estimate runs 3.067 million to 3.749 million.
• The denominator behind the 2024-2025 math is 54,417,758 U.S. residents ages 5 to under 18 as of July 1, 2024.
• Spring 2019 sat near 2.5 million homeschool students. That was the last large pre-pandemic benchmark used in that same synthesis.
• The 2020-2021 school year peaked near 3.721 million. Then it fell to 3.135 million in 2021-2022 after campuses reopened.
• Federal 2018-2019 survey math translated a 2.8% rate into about 1.5 million homeschooled students ages 5-17.
• The 2016 federal wave implied about 1.7 million homeschooled students at a 3.3% rate.
• The 2022-2023 federal parent survey (NCES NHES) weighted 19,562 completed questionnaires to represent about 53 million K-12 students. That is a different universe than the July 2024 Census age count.

Why Do Homeschool Counts Differ Across Surveys?

• The federal household education survey counts a child as homeschooled only if a parent says the student is taught at home for all or some classes. The child also cannot be in a public or private school more than 25 hours a week.
• That federal definition also drops students out for a temporary illness. So short medical absences do not inflate the homeschool rate.
• Census Pulse questions in 2024 did not ask separately about public charter school, hybrid school, or micro-school. As a result, some blended arrangements sit in a gray box.
• Only 29 states plus D.C. publish annual homeschool participation counts. Meanwhile, 21 states still do not run a statewide roll.
• About 11% of the 13 K-12 years fall outside compulsory-attendance ages in some states. So registrars miss 5-, 6-, and 17-year-olds that surveys still treat as school-age.
• West Virginia has been cited at roughly 60% registration compliance. That is a reminder that a state file can be a legal count, not a population count.
• Some families who operate under an umbrella or private-school affidavit never appear on a homeschool register at all.
• The 2024-2025 national synthesis assumed state files undercount by 20% and Pulse undercounts by 15%. Homeschool households are hard to reach and unevenly required to file.
• Pulse reports a percentage, not a national headcount. That is why a 6.5% Pulse reading cannot be multiplied by one denominator and treated as a census.

What Was the Homeschool Share Before COVID?

• The 2018-2019 federal rate was 2.8% of students ages 5-17.
• The same survey put full-time virtual education at 1.2%. Combined “instruction at home” (homeschool plus full-time virtual) sat at 3.7%.
• About 13% of homeschooled students in that wave also participated in full-time virtual education. By contrast, about 1% of students enrolled in a brick-and-mortar school did so.
• Homeschooling ran about 1.7% in 1999, 2.2% in 2003, and 2.9% in 2007.
• The federal rate then flattened: 3.4% in 2012 and 3.3% in 2016. Those readings were statistically indistinguishable from the 2007 figure.
• Pre-pandemic year-to-year growth in the homeschool population was commonly estimated at 2% to 8%. It was not a doubling.
• Rural students in 2018-2019 were homeschooled at 4.7%, against 2.2% to 2.5% in other locales.
• By 2019 the practice was larger than in the 1990s and no longer rising fast. That is the baseline later Pulse spikes get compared against.

Did Homeschooling Drop After Schools Reopened?

• Household Pulse jumped from 5.4% of households with school-age children reporting homeschooling in spring 2020 to 11.1% in October 2020. That is a household share, not a student share. It is the high-water mark most recap charts still use.
• The next federal student survey was fielded January-August 2023 for 2022-2023. It landed at 3.4% homeschooled. That is still above the 2.8% 2018-2019 rate. It is also far below the 2020 household spike.
• Full-time virtual enrollment in that federal file rose from 0.86% of K-12 students in 2018-2019 to 1.8% in 2022-2023. Those students are not automatically counted as homeschoolers.
• Pulse student-share estimates moved from 5.82% in 2022-2023 to 5.92% in 2023-2024. That is a 0.1 point gain rather than a collapse.
• After the 2020-2021 peak, the research synthesis recorded a drop of about 586,000 students into 2021-2022. Then the count climbed back above 3 million.
• Among states that publish rolls, 19 of 21 reported higher homeschool participation in 2024 than in 2023.
• The short version: the emergency spike faded. The pre-2019 share did not return. And the post-reopening dip did not become a round-trip back to 2.8%.

Which States Grew Fastest in 2024-2025?

• Reporting states added homeschool students at an average 4.9% in 2024-2025. That is more than double the old 2% pre-pandemic pace.
• 80% of states that filed 2024-2025 counts showed an increase.
• 36% of those reporting states posted their highest homeschool enrollment on record. Some of those totals beat their pandemic peaks.
• South Carolina led the published growth list at 21.5% over the prior year.
• Vermont rose 17%, Ohio 15%, New Hampshire 14.5%, and Maine 10.8%.
• New Hampshire and Vermont had declined in 2023-2024, then reversed in 2024-2025. So a one-year state rate can hide a two-year sawtooth.
• Louisiana grew only 1%. Even so, it sat on a list of states that have risen every year since the pandemic without a post-reopening dip, alongside North Dakota, South Dakota, and South Carolina.
• Florida’s homeschool count fell 1.7% in 2024-2025.
• Six reporting jurisdictions declined that year: Delaware, the District of Columbia, Florida, Hawaii, New Mexico, and Tennessee.
• 28 states had produced a 2024-2025 homeschool file when that growth tracker was compiled. The reporting lag, plus the 21 non-reporting states, is why a national “growth rate” is still an average of the states that bother to count.

Which States Have the Highest and Lowest Homeschool Shares?

• Pulse readings for 2023-2024 put Alaska at 16.15% of school-age children homeschooled. That was the highest state share in that file, leading all states with highest homeschool rates.
• Tennessee followed at 10.75%, Montana at 9.03%.
• West Virginia sat near 8.23% and Arkansas near 8.16% in the same 2023-2024 Pulse cut.
• A year earlier, Pulse had Alaska at 12.56%, Tennessee at 9.02%, and West Virginia at 8.89%. The top of the table moved, but it did not flatten.
• The low end of that 2023-2024 Pulse file was Connecticut at 3.01%, D.C. at 3.04%, and Massachusetts at 3.39%.
• State registration files tell a different ranking. North Carolina’s published homeschool roll for 2024-2025 was 165,243 students, or 9.47% of that state’s roughly 1.74 million school-age children.
• Massachusetts’ 2024-2025 register showed 12,203 homeschool students, 1.21% of school-age children. Rhode Island showed 3,086, or 2.03%.
• Alaska’s Pulse share rose 3.6 points from 2022-2023 to 2023-2024. That was the largest state gain in that pair of years. D.C. dropped 5.5 points.
• High-share states and high-regulation states are not the same list. New York still added students while sitting in the high-regulation tier. And several no-notice states do not publish a roll at all.

How Does Homeschool Compare With Charter, Private, and Catholic Enrollment?

• A 2026 national sector compilation put traditional public enrollment at 74.0%, charters at 7.2%, tuition-paying private school at 6.1%, magnets at 5.0%, and choice-program seats at 2.8%.
• In broader comparative data on homeschooling vs Catholic and charter schools, Catholic schools enrolled 1.684 million students in 2024-2025.
• Public charter enrollment was 3.784 million in 2023-2024. That is close to the homeschool headcount above, and it is still larger.
• Pulse for 2023-2024 assigned 84.16% of school-age children to public school and 9.92% to private school. That public bucket folds charters and magnets into one number. That is why it does not match the 74.0% “traditional public” cut.
• Pulse private share slipped from 10.06% in 2022-2023 to 9.92% in 2023-2024. At the same time, the homeschool Pulse share ticked up a tenth of a point.
• In a spring 2024 parent survey of 3,191 respondents covering 5,167 K-12 children, weighted student shares came out 73% assigned public, 11% public choice (in-district choice plus charter), 10% private, and 5.7% homeschool.
• 443 of those surveyed children lived in a household that homeschooled at least one sibling, a weighted 7.9%. That is larger than the student-level homeschool rate because families mix sectors.
• Size order depends on the yardstick. Charters still win on published seats. Homeschool wins or ties depending on whether you trust Pulse, a blended synthesis, or the last federal household education survey.

Who Homeschools by Race, Income, and Parent Education?

For complete demographic distributions and socioeconomic tables, see our in-depth research on homeschooling demographics in the United States.

• Pulse respondents with a homeschooled student in 2022-2023 were 60% White, 19% Hispanic or Latino, 12% Black, 2% Asian, and 7% two or more or another race. That is about 40% students of color.
• The older federal household education survey put students of color nearer 29% of the homeschool population. Those are different instruments. They are not a four-point “trend line” you can splice.
• A 2025 parent survey found homeschool parents 69% White against 57% of the U.S. population, 15% Hispanic against 20%, 10% Black against 14%, and 5% Asian against 7%.
• A separate research synthesis has put non-White/non-Hispanic homeschool families at 41%.
• Among Pulse homeschool households in 2023-2024, 9% of responding adults had less than a high school diploma, 29% a diploma or GED, 33% some college or an associate degree, and 29% a bachelor’s degree or higher.
• Income in that same Pulse slice was spread out: 10% under $25,000, 18% from $25,000 to $49,999, 25% from $50,000 to $99,999, 23% from $100,000 to $199,999, 5% at $200,000 or more, and 24% blank.
• The 2019 federal file, a different year and instrument, had about 30% of homeschool families under $50,000, 48% from $50,000 to $99,999, and 22% at $100,000 or more.
• Pulse homeschool respondents clustered in midlife: 38% ages 25-39 and 41% ages 40-54.
• Census Pulse during 2020 recorded Black households reporting homeschooling rising from about 3.3% in spring to 16.1% in the fall. That was the sharpest published racial swing in that emergency year.
• Party ID in the 2025 parent file was 44% Republican and 29% Democratic among homeschool parents, with 44% attending religious services weekly. That is a conservative lean. It is not a single-church census.

How Many Homeschool Students Use Hybrids, Co-ops, or Virtual Classes?

• In the 2025 parent survey, 71% of homeschool students were full-time at home and 29% used another arrangement.
• That 29% split as 18% co-op, 8% hybrid school, and 3% something else.
• Among households with at least one homeschooled child, 44% also used another sector: 35% traditional public, 9% charter, 13% private. Those household percentages can add past 44 because one home can touch two extra sectors.
• A winter 2025 parent poll found 88% of homeschooling parents using some supplemental support, against 44% of parents overall.
• Inside that homeschool group, 44% used online resources, 29% joined a co-op, 27% took an online class, 24% enrolled in an online school, and 10% hired a tutor.
• The 2018-2019 federal survey already showed 13% of homeschoolers in full-time virtual education. That older virtual overlap is not the same thing as a 2025 hybrid-school seat.
• “Homeschool one, homeschool all” fails the household test. Traditional-public households in one mixing study were 82% exclusive to that sector. Homeschool households look more like private and charter households, which also split siblings.

How Long Do Families Stay in Homeschool?

• In the 2023 Cardus adult file, about 36% of U.S. adults who were ever homeschooled did it for only one or two years.
• Fewer than 20% of those adults were homeschooled for all of primary and secondary school.
• A companion adult file put 26% at a single homeschool year and another 17% at two years.
• A research brief using those same adult surveys summarized the short-stay group at roughly 25% for one year. It put the full-career group at 10% to 17%.
• Adults homeschooled for one year spent about 10 years in traditional public schools. Adults homeschooled two or three years spent about 8 years there.
• The one-to-two-year group also averaged more than 3 years in charter or private schools across the rest of K-12.
• Point-in-time shares therefore understate lifetime contact. A 6% snapshot in one school year is not the share of adults who will ever have a homeschool year on their transcript.
• Policy drafts that assume a child who leaves the district roster never returns are writing rules for a minority pattern, not the modal one.

Why Do Parents Choose Homeschooling?

• In federal parent survey data examining why parents choose homeschooling, 83% of homeschool parents named concern about the school environment as a reason.
• Dissatisfaction with academic instruction was cited by 72%.
• A desire to provide moral instruction was cited by 75%, and an emphasis on family life by 72%.
• Religious instruction came in at 53%. That is a majority reason, not the only reason, and not the top-listed one in that wave.
• A nontraditional approach was cited by 50%.
• Special needs were cited by 21%. A physical or mental health problem lasting six months or more was cited by 15%.
• Temporary illness was only 2%. That is why the federal definition tries to keep short medical absences out of the homeschool rate.
• Environment and academics now outrun the older “religion-only” story. Religion is still on the form. It is no longer the whole form.

Do Homeschool Families Use Public Funds or ESAs?

• In the winter 2025 EdChoice parent poll of homeschoolers, 22% said they had access to public funding. Another 25% did not know. And more than half said they had none.
• Among the small group that reported access, 63% said they used the money.
• Among those without access, 71% said they would use public funds if they could. Combined, 68% of that homeschool sample either use public money or would.
• Arizona’s ESA roster crossed 100,000 students by January 20, 2026. The state then reported 104,644 students on 2026-27 contracts as of August 31, 2026.
• The same program had about 12,000 students in 2021-22 and about 88,000 in 2024-25. That is a 633% rise after universal eligibility, equal to about 7% of Arizona’s school-age population in 2024-25.
• Typical 2025-2026 Arizona awards for general-education students ran about $7,000 to $8,000. Signing the contract replaces a county homeschool affidavit. So an ESA student is often missing from the homeschool register on purpose.
• An Arizona spend analysis found about 60% of fall 2024 ESA dollars, roughly $120 million, going to private-school tuition. The rest spread into curriculum, tutoring, and other services homeschool-style families actually buy.
• Choice programs of all types enrolled about 1.5 million students in 2025-2026 across 34 states plus D.C. and Puerto Rico. 21 ESA programs sat in 18 states as of mid-2026.
• Public money and “homeschool” are no longer synonyms for unpaid parent labor. In affidavit states they still can be. In ESA states the count and the legal label diverge.

How Much Does Homeschooling Cost Families and Taxpayers?

• A 2024 poll of more than 4,000 homeschool families put average out-of-pocket spending at $1,295 per elementary student and $1,636 per middle or high school student.
• That same poll put a two-child household near $3,104 a year in combined student spending.
• A long-running research synthesis still quotes about $600 per student per year. Treat that as a different, thinner estimate, not an update of the 2024 poll.
• Community budget bands used in 2026 guides run from under $200 on library-and-used-book years to $1,500-$3,000+ for a boxed program plus paid classes.
• Public schools spent an average $18,853 per pupil in 2025, plus capital outlays, in the teacher-association compilation used by that research synthesis.
• Multiply that per-pupil public figure by the 2024-2025 homeschool headcount and the implied taxpayer outlay not spent is more than $64 billion. That is an avoided-cost identity. It is not cash mailed back to districts.
• There is still no federal income-tax deduction for homeschool curriculum. The educator-expense deduction rises to $350 for tax year 2026 and still excludes home instruction.
• Nine states offered an income-tax credit or deduction a homeschooler could actually claim as of August 2026. The range ran from Ohio’s $250 per student to Idaho’s refundable $5,000 (and $7,500 where a qualifying disability requires extra staff).
• Florida’s homeschool-facing Personalized Education Program has been quoted around $8,000 per student, with county-level bands from about $7,613 to $11,950.
• Writers disagree on 529 plans. One 2026 tax read says a 2025 federal expansion dropped homeschool language before enactment. Another says K-12 withdrawals rose to $20,000 per student on January 1, 2026, and can cover curriculum if the state conforms. Check the statute in force before you treat either line as settled.

How Do Homeschooled Students Score on Tests?

• A 2026 systematic review by NHERI researchers found 62% of peer-reviewed achievement studies showing homeschool students statistically ahead of students in institutional schools.
• The usual cited gap on nationally normed achievement tests is 15 to 25 percentile points above the public-school mean. That public-school mean sits near the 50th percentile.
• A 2026 nationwide achievement study reported homeschool means between the 75th and 79th percentiles in reading, language, math, social studies, and science.
• After weighting for free-or-reduced-price lunch eligibility and race or ethnicity, that study’s homeschool means sat between the 74th and 82nd percentiles. That is about 0.75 to 1 standard deviation above the test norms.
• Subject gaps versus the norm group in that sample were about 29 percentile points in reading, 26 in social studies, 27 in science, and 25 in language and in math.
• A 2015 analysis of Black homeschool students found them 23 to 42 percentile points above Black public-school students on the same class of tests.
• Those files also report scores above average across parent education and household income bands. They also find little relationship between a parent’s teacher certification and the child’s test score.
• Degree of state regulation has not shown a stable link to achievement in that research line. That is a correlational claim. It is not proof that paperwork is free.
• SAT and ACT averages for homeschool cohorts have historically run above the national mean. But the SAT publisher no longer publishes a separate homeschool cut each year. So treat older 60-100 point SAT gaps as historical, not a 2026 release.
• Selection is the standing caveat. Families who sit for a nationally normed test, and who send scores to researchers, are not a random draw of every affidavit household.

How Do States Regulate Homeschooling?

• Homeschooling is legal in all 50 states and D.C. The fight is over notice, testing, and who holds the file.
• The usual legal map uses four tiers: no notice, low (notice only), moderate (notice plus a test or evaluation), and high (notice, testing or evaluation, plus curriculum review or home visits).
• Ten jurisdictions are commonly listed as no-notice: Connecticut, Idaho, Illinois, Indiana, Missouri, New Jersey, Oklahoma, South Carolina, Texas, and Wyoming. Alaska, Florida, Iowa, and Michigan sometimes appear on that list only for their least-regulated legal route.
• High-regulation lists almost always include Massachusetts, New York, Pennsylvania, Rhode Island, and Vermont. Some maps add North Dakota or D.C.
• 29 states plus D.C. publish an annual homeschool count. The other 21 do not. That is why national tallies keep leaning on Pulse and blended estimates.
• One 2024 policy scan counted 31 states requiring some form of registration, documentation, or notice.
• A child-protection review has counted 19 states that do not require core subjects such as English, science, or math to be taught under the homeschool statute.
• Where test floors exist they are low. Colorado’s line is the 13th percentile, and two years under it can end home instruction. New York wants a composite above the 33rd percentile or a year of growth. Virginia and West Virginia use the 4th stanine, about the 23rd percentile.
• Minnesota, North Dakota, and Iowa use the 30th percentile as a trigger for evaluation or a remediation plan. Oregon’s line is the 15th percentile.
• New York’s high-regulation packet still includes a letter of intent within 14 days, an Individualized Home Instruction Plan within 4 weeks, quarterly reports, and tests in grades 4, 6, and 8.
• Pennsylvania still stacks an annual affidavit, quarterly portfolios, an evaluator letter, and tests in grades 3, 5, and 8.
• Strict statutes have not frozen enrollment. New York is high regulation and still growing. Several no-notice states do not publish a roll. So “easy law, high share” is not a table you can fill for all 50.

Recent Developments

• North Carolina’s homeschool register fell to 152,897 students in 2025-2026, down 7.4% from 165,243 the year before, after two years of gains.
• North Carolina private schools moved the other way: 143,998 students in 2025-2026, up 8,260 or 6.1%. The state’s private-school voucher program kept expanding. So homeschool and voucher-funded private school are now substitutes in at least one large reporting state.
• New York State counted 53,905 homeschool students in the latest published recap, more than triple the 2009-2010 figure, with more than 14,000 in New York City. That is a high-regulation state adding students anyway.
• New York’s public per-pupil spend is cited at about $37,000. Local critics keep pairing that figure with the homeschool rise.
• Arizona’s ESA roster was 104,644 students on 2026-27 contracts as of August 31, 2026, after crossing 100,000 in January. In September 2026, Apache Junction Unified began selling à la carte public classes, lunch, and sports to ESA students.
• Choice programs nationwide enrolled about 1.5 million students in 2025-2026. Mid-2026 tallies still counted 21 ESA programs in 18 states.
• New Jersey proposals that would have ended the state’s no-notice status (intent letters, portfolio talk, wellness-check language) missed the January 13, 2026, deadline. They can be refiled.
• In spring 2026, Pennsylvania inspectors treated a homeschool hybrid as an unlicensed child-care site and issued a cease-and-desist. That is a live fight over whether a hybrid is a school, a homeschool, or day care.
• A July 4, 2025, federal package created a national voucher-style program limited to states that opt in. It also reopened the 529 K-12 fight. Tax guides still disagree on whether homeschool curriculum is in or out.
• The federal household education survey that produced the 3.4% 2022-2023 rate has been halted. The future of education items on Census Pulse was already being described as uncertain in early 2025. So the next “official” national percentage may arrive late, or not at all.
• Winter 2025 parent fieldwork on supports and public-fund access was released in March-April 2026. A 2026 peer-reviewed achievement paper filled a 15-year gap in nationwide homeschool test-score samples.
• Texas began running a large new voucher pot, reported near $1 billion. That changes who can afford a homeschool year in that state even if the homeschool statute itself stays loose.

Editor’s Choice

• Carry two national rates, not one blended fake average. The last federal student survey is 3.4% (2022-2023). The 2024-2025 working synthesis is 6.262%.
• If you need a headcount for 2024-2025, use 3.408 million and keep the band 3.067 million to 3.749 million. Do not back-solve Pulse’s 6.530% against a single national denominator.
• The 2020 chart that still circulates is a household number: 11.1% of households with school-age children in October 2020, up from 5.4% that spring. It is not the student share.
• Among states that file rolls, 2024-2025 growth averaged 4.9%, and 80% of those states were still up. South Carolina’s 21.5% jump is the ceiling, not the mean.
• Geography swamps the national average. Alaska’s 2023-2024 Pulse share was 16.15%. Connecticut’s was 3.01%.
• Watch the legal label. Arizona’s ESA file had 104,644 students on 2026-27 contracts by August 31, 2026. Many of those students drop off the homeschool affidavit roll on purpose.
• North Carolina is the warning case for substitution. Homeschool registration fell 7.4% to 152,897 in 2025-2026 while private enrollment rose 6.1%.
• Most homeschool students are not a closed household. 71% were full-time at home in the 2025 parent file. And 44% of homeschool households also touched another sector.
• Duration wrecks permanence assumptions. About 36% of adults who were ever homeschooled did it for one or two years. Under 20% did K-12 at home.
• Money is split-screen. Families in a 2024 poll spent $1,295 (elementary) to $1,636 (secondary) per student. Public per-pupil spending sat at $18,853 in 2025.
• Only 22% of homeschool parents in the winter 2025 poll said they currently have public funds. 68% either use them or would.
• Environment, not a single religious line, leads the motive file. 83% cited school environment in 2022-2023, 72% academics, and 53% religious instruction.
• The next clean federal percentage is at risk. The household education survey behind the 3.4% figure has been halted. Pulse education items were already in doubt in 2025.

Frequently Asked Questions

Conclusion

Use the cut that matches the decision. A district planner should track notices and ESA exits separately. A falling homeschool register can sit next to rising home-based instruction paid from a state wallet. A parent comparing “the” national percentage should keep 3.4% and 6% on the same page and refuse a homemade average. Anyone writing a rule that treats homeschool as a 13-year exit is legislating against the one-to-two-year pattern that shows up in adult surveys. The share is larger than it was in 2019. It is smaller than the 2020 household spike.