New Business Applications by U.S. State: 2025 Ranking

Overview: what a business application means

The Census Bureau's BA series measures selected EIN applications made through IRS Form SS-4. It excludes filings related to trusts, estates, tax liens, certain financial transactions and specified other categories. The series is distinct from high-propensity applications (HBA), projected employer formations and firms that actually begin paying employees.

By raw number of filings, five states exceeded 250,000 in 2025: Florida, California, Texas, New York and Georgia. A count-based ranking is useful for comparing filing volumes, but it is not adjusted for population, the number of existing firms or whether the applicant will establish a payroll business. Registration geography should not be assumed to equal where a firm ultimately operates.

The 16-state South represented 44.23% of the 50-state application total, compared with 24.55% for the 13-state West. These are shares of the 5,656,891 applications in the 50-state comparison, not shares of all EIN filings or all firms established.

Top 10 states by new business applications in 2025

The three leaders together accounted for 1,745,811 applications. The tenth-ranked state, New Jersey, recorded 161,138, just 137 fewer than ninth-ranked Pennsylvania.

Top 10 · Census BFS BA · 2025 calendar-year NSA totals

Each Top 10 value is a source-published 2025 BA (NSA) annual count, classified industry_observed_value. Rank and Top 3 concentration are StatRanker calculations. The annual counts were checked against two publishers; the entire 50-state monthly data file was not independently reprocessed.

Chart: Top 20 states by 2025 applications

Bars compare absolute totals, with Florida set to 100% of the leader's count. The labels and values match the full table; bar lengths are calculated as each state's applications divided by Florida's 647,734.

Scale: BA count relative to Florida = (state applications / 647,734) × 100. Bars are a visual comparison only; original values appear beside each bar.

Methodology and data quality

Reconciliation: StatRanker summed the 50 published state annual counts to obtain 5,656,891. The D.C. figure of 14,945 was checked separately against its own monthly Census series distributed through FRED, rather than inferred by subtraction. Thus 5,656,891 + 14,945 = 5,671,836, matching the separately published national annual count. Share calculations on this page use 5,656,891 as denominator and do not claim a share of all possible EIN filings.

StatRanker-calculated indicators (not source-published annual figures): 50-state total = sum of 50 reported BA values = 5,656,891; mean = 5,656,891 / 50 = 113,137.82; median = (Utah, rank 25: 72,787 + Wisconsin, rank 26: 71,475) / 2 = 72,131. Top-N concentration = sum of the first N ranked states / 5,656,891. Region share = regional BA sum / 5,656,891; state share = state BA / 5,656,891. A rank gap = larger reported count minus smaller reported count. The bottom-21 comparison uses ranks 30–50. These are calculated_value descriptive statistics from confirmed 2025 observed annual inputs, not Census forecasts or projections. No imputation or inflation/population adjustment was applied; percentages are rounded after calculation.

Reproduction protocol and verification limit: filter the Census monthly file to series=BA_BA, sa=U, naics_sector=TOTAL, year=2025 and each state geo. For each state, annual BA = sum(jan, feb, mar, apr, may, jun, jul, aug, sep, oct, nov, dec). Each monthly input is an observed Census BA count; every included row covers all twelve months of 2025. This is the reproducible method, not a claim that StatRanker imported and recalculated all 600 state-month cells. For Florida, the twelve monthly values verified via FRED were 52,965, 53,408, 57,940, 71,550, 49,518, 52,521, 59,351, 52,324, 49,751, 61,641, 44,214 and 42,551, summing to 647,734. Direct monthly checks were also performed for California, Alabama and D.C.

Limitations: BA counts do not establish the number of companies that opened, hired, survived, registered as LLCs or generated revenue. They are not application rates per person. The state recorded for an application is not necessarily the state where a firm ultimately conducts operations. No causal explanation of high rankings is inferred from these data.

Full ranking: all 50 U.S. states

Search states, filter by Census region or value status, and choose a sort or Top N view. Top 10 and Top 20 always refer to the original count-based ranks, even if the display order changes.

Showing 50 of 50 states

For every row: 2025 observed application count, unit applications, original Census BFS BA non-seasonally-adjusted measure; numeric publishers Mainstay Filing and SMB Statistics; status industry_observed_value. Counts are source-published annual aggregates, not independent StatRanker monthly calculations. The state total, ranks, median, mean, concentration and other analytical measures are calculated from those observed annual numbers.

Four findings from the 2025 distribution

What the 2025 filing ranking can tell us

Filing volume is not startup probability. Florida recorded 647,734 qualifying EIN applications (rank 1), while Wyoming recorded 79,023 (rank 22). Comparing these absolute totals does not tell us which state's residents are more likely to launch a company; that question needs a consistently defined population denominator and care about where EIN paperwork is filed.

A filing is not an employer. A qualifying BA application might never result in a payroll firm. The separate Census high-propensity and actual employer-formation series answer different questions and cannot be substituted into this table.

Regional shares have a narrow denominator. The South's 44.23% refers to 2,501,935 of the 5,656,891 applications across 50 states. It does not imply that the South produced 44.23% of new operating businesses, incorporated firms or jobs. None of these counts alone establish why one state ranks above another.

Frequently asked questions

Sources and roles