Compare U.S. cities with the largest population losses in 2024–2025. Read exact Census changes, relative declines, historical context and source notes.
New York leads this numeric population-loss ranking with 12,196 residents lost. The comparison covers July 1, 2024 to July 1, 2025.
New York, New York had the largest numeric decline: 12,196 residents between July 1, 2024 and July 1, 2025. That is 0.14% of its starting population. Memphis, Tennessee lost 4,575 residents over the same interval.
8,789 of the 19,471 eligible cities declined, while 8,484 increased and 2,198 were unchanged. This page orders only the declining records by residents lost. It does not present the lowest growth rate as though it were the largest numeric loss.
Numeric losses and relative declines produce different leaders
Among declining cities, Aurora, New York had the largest proportional decrease at 46.94%. Its numeric change was 291 fewer residents. No minimum-population threshold is applied to this numeric-loss ranking; a small place can appear in the comparison of proportional losses.
The reported change compares two estimates from the same vintage. A revision to the 2024 estimate in a later publication is a separate event from the 2024–2025 decline. Differences between website figures should be checked against observation dates and source vintages before they are interpreted as population loss.
Comparison universe: Active whole incorporated-place records (SUMLEV 162); consolidated-government balances are excluded. Source: PEP Vintage 2025. Population date: July 1, 2025. Growth interval: July 1, 2024–2025. Ties share a competition rank; subsequent ranks skip tied positions. Filters produce a subset of this universe and do not change the national profile rank.
Chula VistaCalifornia · Incorporated place · 0613392
275,533
-1,214
-0.44%
5,537.9
Page 1
How the recent decline compares with the longer series
For New York, the five-year change from July 2020 to July 2025 was -166,559 residents, or -1.90%. A single declining year can occur inside a period of overall growth, and a longer decline can contain individual growing years. Read the full dated series on the profile.
The city source does not publish city-level migration or natural-change components. The decline therefore cannot be attributed to births, deaths or migration using this totals file alone. Neither a decline nor a growth rank by itself establishes a policy effect or a reason people moved.
Sources, calculation rules and downloadable results
Source: sub-est2025.csv, Census Population Estimates Program, Vintage 2025, variables POPESTIMATE2024 and POPESTIMATE2025. Density additionally uses 2025 Gazetteer ALAND joined by official code. The retained source snapshot is pep-2025-r1; calculations use unrounded values.
Change order uses 2025 population minus 2024 population. Equal unrounded metric values share a competition rank, and the following rank skips the tied positions. Display rounding can make different underlying values look identical; the table order uses the originals.
Export visible rows after filtering to preserve the displayed subset. Use complete dataset downloads for all observations, or population research tools for a longer comparison. A ranking is a dated descriptive result, not an official forecast. Publication vintages can revise earlier years; retain the release identifier with your research.