Saint Paul Crime by Neighborhood (2026)

Reported crime incidents for each of Saint Paul's 17 district council neighborhoods over the past 12 months, updated weekly from the City of Saint Paul's official dataset. Click a neighborhood for monthly trends, top incident types, and a live map.

Neighborhood	Safety Score	Weighted incidents / sq mi	Past 30 days	Past 12 months
Conway / Battle Creek / Highwood	84	59	70	1,053
Como	75	79	25	489
St. Anthony Park	72	86	22	463
Highland	69	96	65	1,115
West Side	67	100	34	791
Summit Hill	64	108	12	262
Macalester–Groveland	62	112	22	861
West Seventh	56	128	37	571
Greater East Side	48	149	63	1,236
Hamline–Midway	32	234	72	941
North End	29	262	81	1,360
Payne–Phalen	29	260	97	1,714
Dayton's Bluff	25	287	65	1,039
Union Park	21	309	116	1,922
Frogtown (Thomas–Dale)	19	322	54	803
Summit–University	14	353	79	1,312
Downtown (Capitol River)	1	776	96	1,661

How these scores work

The score is a citywide rank, not a grade. Each district's incidents over the past 90 days are weighted by how strongly they signal public-space danger, divided by the district's land area, and ranked against the rest of Saint Paul. Higher = quieter. The "weighted incidents per square mile" column is the exact number being ranked — the ordering follows it.
Not every incident counts equally. Shots fired and robbery carry the most weight; theft carries little; domestic incidents and narcotics cases carry the least, because they say more about relationships and police patrols than about the street outside.
Commercial corridors count in full. Districts containing major retail nodes — the Midway, Grand Avenue, downtown — score lower than their residential blocks feel, because shopping districts concentrate incidents. The reverse also applies: districts with large parks, industrial land, or river flats spread their incidents over more area.
For where you actually live, use the map. Saving an address on the dashboard scores just the area around it — within the same district, a residential street and a shopping district can be 80 points apart.

Raw 30-day and 12-month counts are not adjusted for district size or population. Full methodology, including every weight →