Which US cities got safer — and which got worse

Almost every crime ranking is a single-year snapshot. This one is a direction: the change in total reported crime rate for 661 US cities between 2021 and 2025.

531

cities improved

107

cities got worse

essentially flat

−20.3%

median change

Biggest improvements, 2021–2025

Largest percentage drops in total reported crime rate.

#	City	Population	Change in crime rate
1	Middletown, OH	52,613	−69.6%
2	West Haven, CT	55,336	−62.7%
3	Encinitas, CA	60,923	−60.8%
4	East Hartford, CT	51,190	−58.4%
5	Sammamish, WA	66,235	−57.2%
6	Fayetteville, AR	105,335	−57.1%
7	San Marcos, CA	97,052	−56.0%
8	Newark, OH	51,776	−55.5%
9	Kent, WA	136,591	−54.7%
10	St Charles, MO	72,881	−53.3%
11	Vista, CA	99,445	−53.2%
12	Turlock, CA	72,919	−52.8%
13	Sarasota, FL	58,428	−52.7%
14	South San Francisco, CA	64,363	−52.3%
15	Murray, UT	50,895	−52.0%
16	Burleson, TX	58,540	−51.3%
17	San Francisco, CA	815,330	−50.7%
18	Sandy, UT	91,841	−50.7%
19	St Peters, MO	62,360	−50.7%
20	College Station, TX	129,902	−50.6%
21	Broomfield, CO	79,451	−50.1%
22	Bristol, CT	62,563	−50.0%
23	North Port, FL	97,818	−49.7%
24	Roswell, GA	92,178	−49.7%
25	Indio, CA	95,514	−49.4%

Biggest deteriorations, 2021–2025

Largest percentage rises in total reported crime rate.

#	City	Population	Change in crime rate
1	Aliso Viejo, CA	50,060	+82.3%
2	Bowie, MD	58,471	+79.7%
3	Waltham, MA	66,089	+64.5%
4	Valdosta, GA	55,729	+64.0%
5	Clifton, NJ	90,406	+60.7%
6	Harrisburg, PA	50,789	+56.0%
7	Gary, IN	67,197	+54.3%
8	Bolingbrook, IL	74,990	+47.3%
9	Allentown, PA	127,494	+44.7%
10	Baldwin Park, CA	68,586	+43.3%
11	Bayonne, NJ	75,204	+42.6%
12	Bethlehem, PA	80,424	+42.1%
13	Jersey City, NJ	305,249	+40.5%
14	Upland, CA	79,844	+35.9%
15	Somerville, MA	82,255	+35.1%
16	Cambridge, MA	121,634	+34.9%
17	Miami, FL	498,792	+34.5%
18	Sterling Heights, MI	134,402	+32.8%
19	Lawton, OK	89,942	+31.6%
20	Normal, IL	53,374	+31.5%
21	Lake Forest, CA	87,471	+31.0%
22	Mount Prospect, IL	55,157	+26.5%
23	West Sacramento, CA	56,549	+26.5%
24	West New York, NJ	53,036	+26.3%
25	Tigard, OR	57,885	+24.7%

How this is calculated — and what it can't tell you

Change is measured between a city's first and last reporting year in our series, using total reported offenses per 100,000 residents.
Cities below 50,000 residents are excluded. At small populations a handful of incidents moves the rate by double digits, and the "trend" is noise rather than signal.
Cities with fewer than 3 reporting years, or flagged for thin agency coverage, are excluded entirely.
Movements under 2% either way are reported as flat, not as a direction.
This measures reported crime. A change in reporting behaviour or in an agency's participation can move these numbers without underlying crime changing — which is why the excluded-city rules above are strict.

Full method on the methodology page, and the underlying series is in the free data download.