How big should your emergency fund be? A workable answer is your essential monthly expenses, minus any income that would continue if you lost a paycheck, multiplied by the months you want to cover, plus a one-time reserve for deductibles. The familiar "3 to 6 months" is a convention for that multiplier rather than a rule, and your own situation can justify fewer months or considerably more.
Key takeaways
- Emergency fund sizing starts with essential spending (housing, food, utilities, insurance, minimum debt payments, transportation), not with total spending or income.
- Income that would keep coming in, such as a partner's pay, reduces the monthly gap you need to fund; unstable or single-source income raises the number of months.
- In October 2026 the median unemployed person had been unemployed for 11.9 weeks, but 27.6% had been unemployed for 27 weeks or more, so a six-month cushion is a reasonable planning case, not an extreme one.
- State unemployment benefits are capped well below many paychecks (as of July 2026, a maximum of $275 a week in Alabama versus $869 in New York), and self-employment income generally does not count toward a regular claim, so the fund should not assume benefits replace most of your pay.
- The Federal Reserve found that 55% of adults in 2025 said they had three months of expenses set aside, and 12% could not cover a $400 expense by any means, so a partial fund is common and still useful.
The sizing formula in plain terms
Start with the bills that continue whether or not you have a job: rent or mortgage, utilities, groceries, insurance, transportation, minimum debt payments, and childcare if you could not pause it. Leave out spending you would stop in a crisis, such as dining out, subscriptions and travel. Essential spending is almost always lower than total spending, so a fund sized on total spending overshoots.
Target = (E - C) x M + D
- E - essential monthly expenses
- C - monthly income that would continue after a job loss (a partner's take-home pay, steady side income); zero for a single earner
- M - months of cover, chosen from the factors below
- D - one-time reserve for known large out-of-pocket costs, such as a health plan deductible or a home insurance deductible
The formula is only as good as the estimate of E. Three months of bank and card statements give a starting estimate; add essential bills paid once or twice a year, such as insurance premiums or car registration, as a monthly average.
Why 3 to 6 months is a convention
The Federal Reserve's household survey uses three months of expenses as "one common measure of financial resiliency," and in 2025, 55% of adults said they had set aside money for that long, down from a high of 59% in 2021 (Federal Reserve, Economic Well-Being of U.S. Households in 2025). That makes three months a research yardstick, not a regulatory requirement or a tested optimum.
The Consumer Financial Protection Bureau's guide to building an emergency fund does not give a number of months. It says the amount "depends on your situation," suggests looking at the most common unexpected expenses you have had in the past and what they cost, and adds that even a small amount can provide some financial security. That supports a staged approach: a small first buffer, then a target sized from the formula.
The Fed's own data show why a fixed number misleads. The share of adults who would cover a hypothetical $400 expense with cash or its equivalent was 63%, while 12% said they could not pay it by any means. A household in the second group and a household with $60,000 in savings need very different advice, and "3 to 6 months" gives both the same sentence.
Factors that move your target
Each factor changes E, C, M or D. The table shows the direction; the size of the effect is a judgment call.
| Factor | Pushes the target | Why |
|---|---|---|
| Income is steady and in demand (salaried, one employer) | Down (fewer months) | Shorter expected gap between jobs |
| Income is variable, commission-based or self-employed | Up (more months) | Lean months can occur without a job loss, and unemployment benefits may not apply |
| Second earner with separate employer or industry | Down (higher C) | Only part of household income stops at once |
| One earner supports the household | Up (C = 0) | The whole essential budget has to be funded |
| Children or other dependents | Up (higher E) | Essential costs are larger and harder to cut |
| Own a home | Up (higher D) | Repairs and deductibles fall on you rather than a landlord |
| High-deductible health plan | Up (higher D) | You pay most covered costs yourself until the deductible is met |
| Long job search expected (specialized field, small local market) | Up (more months) | See the duration data below |
The health plan piece
For D, the plan's deductible is the usual starting point, with the out-of-pocket maximum as the ceiling for a bad year. For 2026, Marketplace plans cap in-network out-of-pocket costs at $10,600 for an individual and $21,200 for a family, and monthly premiums and out-of-network charges do not count toward that cap (HealthCare.gov). Employer and other plans state their own maximum in the plan documents, so use the number for your plan. Funding the full maximum in cash is out of reach for many households; the deductible is the more practical number to set aside, and the maximum shows the worst case.
Job search and benefit limits
Two public data sets frame how many months to plan for, though neither measures your own situation.
How long unemployment lasts
In the Bureau of Labor Statistics' Employment Situation table A-12 for October 2026 (seasonally adjusted), the median duration of unemployment was 11.9 weeks, and the mean was 27.1 weeks. Of the roughly 7.1 million unemployed people in the table's duration breakdown, about 2.2 million (30.5%) had been unemployed under 5 weeks, 1.7 million (24.4%) for 5 to 14 weeks, 1.2 million (17.5%) for 15 to 26 weeks, and 2.0 million (27.6%) for 27 weeks or more.
Converted to months, 11.9 weeks is about 2.75 months and 27.1 weeks is about 6.25 months. The gap between the median and the mean shows a long tail: a few long spells pull the average up. Treat these figures with care, though. The BLS defines duration as how long people have been continuously looking, and notes that it measures spells still in progress; it is not how long it takes to find a job. They are a gauge of risk, not a forecast of one person's search.
What unemployment benefits replace
Each state sets its own eligibility rules; in general, the Department of Labor says you usually qualify if you are unemployed through no fault of your own and meet work, wage and any additional state requirements. The amounts differ widely. In the Department of Labor's Significant Provisions of State Unemployment Insurance Laws (effective July 2026), the maximum weekly benefit amount is $275 in Alabama and Florida, $450 in California, $605 in Texas and $869 in New York. The number of benefit weeks also varies: New York pays a uniform 26 weeks, Arkansas lists 9 to 12, and Florida lists 9 to 12 weeks, with Florida's count depending on base-period wages and the state unemployment rate (the document notes that some states pay extra weeks in limited circumstances, such as high unemployment). Self-employed workers should not plan on benefits: the Department of Labor says income from independent-contractor work is not considered for a regular claim's work and wage requirements, although anyone can apply and the state decides eligibility.
Because each state's benefit is capped, the share of a paycheck it replaces falls as pay rises. For a worker who earned $1,200 a week before the job loss, the caps above work out to at most 22.9% of that wage in Alabama, 37.5% in California and 50.4% in Texas (computed here from the state maximums; the real benefit also depends on base-period wages). For a conservative fund, the sound approach is to treat benefits as a possible bonus to the runway and not subtract them from the gap, unless you have checked your state's formula against your own earnings.
Three illustrative households
These households are hypothetical; every dollar input is an assumption, so swap in your own.
| Single renter, steady salaried job | Two earners, two children, mortgage | One earner, two children, variable pay | |
|---|---|---|---|
| Essential monthly expenses (E) | $2,800 | $6,500 | $5,200 |
| Continuing income if one job stops (C) | $0 | $3,600 (other earner's take-home) | $0 |
| Monthly gap (E - C) | $2,800 | $2,900 | $5,200 |
| Months chosen (M) | 3 | 4 | 6 |
| One-time reserve (D) | $1,500 (health deductible) | $5,500 ($3,000 health plus $2,500 home insurance deductible) | $3,000 (health deductible) |
| Target | $9,900 | $17,100 | $34,200 |
The arithmetic: $2,800 x 3 + $1,500 = $9,900; $2,900 x 4 + $5,500 = $17,100; $5,200 x 6 + $3,000 = $34,200. The months are judgment calls guided by the factor table: the steady salaried worker gets 3; the two-earner family gets 4, one more than the single renter because children and a home make a long gap harder to absorb, while the continuing income is already counted in C; the one-earner household with variable pay gets 6 months, about 26 weeks, close to the 27.1-week mean duration in the BLS data.
Two points stand out. First, the two-earner household needs $7,200 more than the single renter even though its monthly gap is almost the same ($2,900 versus $2,800): $3,200 of the difference comes from the extra month and $4,000 from larger deductibles. Second, the answer is sensitive to what the second income does. If both earners worked in the same industry and both lost income, the gap becomes the full $6,500 and the four-month target rises to $6,500 x 4 + $5,500 = $31,500. Households with two earners at the same employer or in the same industry may want to size against that case.
A partial balance helps while it grows. The Fed survey found that 86% of adults who always had money left over at the end of the month said they had three months of savings, compared with 13% of those who never did, a sign that monthly budget room and savings go together.
This article is general education, not personal financial advice. A licensed financial professional can review your own budget, taxes and risks.
Frequently asked questions
Does a regulator set the 3 to 6 month rule?
No. The Federal Reserve's survey uses three months of expenses as a common measure of resilience, and the CFPB's emergency fund guide says the right amount depends on your situation. The months figure is a convention.
Can a credit card limit count toward the fund?
No. A limit is borrowing capacity, and the Fed's survey separates paying with cash, savings or a card paid off at the next statement from carrying a balance, borrowing or selling something. A card used for an emergency and then carried at interest turns the emergency into debt, which does not shrink the essential bills the fund is meant to cover.
Is $400 a good first target?
The $400 figure is a survey question the Federal Reserve has asked since 2013 to track how households handle a small surprise cost, not a recommended balance. The Fed notes that in 2022 it asked one-fifth of respondents about $500 instead, and the share paying in cash changed by only 0.5 percentage points. A small first balance is reasonable because the CFPB says even a small amount can provide some financial security, and you can then raise it toward the formula's target.
Do retirement savings or other assets count?
That is a personal choice with trade-offs the Fed does not resolve. Among people without a dedicated rainy-day fund, the survey found that some could still cover three months by borrowing, selling assets or drawing on other savings (15% of all adults), and it notes that some people would tap assets such as retirement funds after three months without a job that they would not tap for a smaller emergency today. Funds not intended for this purpose are best kept out of the formula unless you have checked the costs of withdrawing them.
How often should the target be recalculated?
Recalculate when E, C, M or D changes: a new lease or mortgage payment, a change in household size, a second earner starting or stopping work, a new health plan, or a move to variable income. Otherwise, an annual look at last year's essential spending catches gradual drift.
Setting your own number
- List essential monthly expenses (E) from the last three months of statements, leaving out spending you would stop.
- Estimate continuing income (C) conservatively; use zero if the same event could affect every earner.
- Pick M by checking the factor table: 3 for a steady, in-demand job, 6 where income is variable or one earner supports the household, and more if a long search is likely in your field.
- Add D from your plan's deductible and your homeowner's or renter's deductible.
- Compute (E - C) x M + D, then set a first milestone you can reach soon, and revisit the target when your situation changes.
Sources
- Federal Reserve Board, Economic Well-Being of U.S. Households in 2025 (May 2026)
- Consumer Financial Protection Bureau, An essential guide to building an emergency fund
- Bureau of Labor Statistics, Employment Situation, Table A-12 (October 2026)
- Bureau of Labor Statistics, Current Population Survey definitions
- U.S. Department of Labor, Significant Provisions of State Unemployment Insurance Laws (effective July 2026)
- U.S. Department of Labor, Unemployment insurance
- U.S. Department of Labor, Unemployment.gov: benefits for independent contractors and the self-employed
- HealthCare.gov, Out-of-pocket maximum/limit
