Scores are global-scale research estimates. R marks work in progress; none are travel advice.
Found a hotel? Check the area around it before booking.
Search sends your query and selected city to OpenStreetMap Nominatim. Don't enter private addresses or sensitive information. Search policy
Know before you roam
Compare Prague areas on the map, with local evidence and source details.
Safrom's citywide research estimate is 87/100. Higher scores mean safer on the model's scale. Evidence as of . Confidence: C.
57 areas with estimates. Area evidence as of .
smoothed_injury_fatal_share_per_100_recorded_crashes · 2025-01-01/2025-12-31.
2025 police-recorded road crashes, spatially joined to 57 Prague municipal districts. The local metric is the share involving injury or death, stabilized with 20 crash equivalents at the city mean. It is not crime or individual crash probability. 11,029 crashes join exactly one municipal district, including 2,095 injury/fatal crashes. The police Prague jurisdiction has 11,014 crashes and is not identical to the municipal boundary; four jurisdiction-coded points do not join. No points are snapped to a district. Sparse districts are shrunk toward the city injury share; raw aggregate counts are retained. Reporting rules changed during 2025, including the property-damage reporting threshold. Compare only this same-period recorded-crash series. IPR municipal boundaries retrieved 2026-10-27. Topology-preserving simplification with area drift below 0.5%. Bruce explicitly approved limited aggregate use on 2026-10-27: publish self-calculated area aggregates and scores with official attribution; do not republish incident records, report prose or images. This is a project source-admission decision, not a claim of an open license. The citywide score uses Czechia national priors. Local evidence adjusts only mobility.
Download the scored safety layer and calculation evidence (GeoJSON)
2025 police-recorded road crashes, spatially joined to 57 Prague municipal districts. The local metric is the share involving injury or death, stabilized with 20 crash equivalents at the city mean. It is not crime or individual crash probability. 11,029 crashes join exactly one municipal district, including 2,095 injury/fatal crashes. The police Prague jurisdiction has 11,014 crashes and is not identical to the municipal boundary; four jurisdiction-coded points do not join. No points are snapped to a district. Sparse districts are shrunk toward the city injury share; raw aggregate counts are retained. Reporting rules changed during 2025, including the property-damage reporting threshold. Compare only this same-period recorded-crash series. IPR municipal boundaries retrieved 2026-10-27. Topology-preserving simplification with area drift below 0.5%. Bruce explicitly approved limited aggregate use on 2026-10-27: publish self-calculated area aggregates and scores with official attribution; do not republish incident records, report prose or images. This is a project source-admission decision, not a claim of an open license. The citywide score uses Czechia national priors. Local evidence adjusts only mobility.
Boundaries: IPR Praha · Městské části Prahy (CC BY; datový podklad © IPR Praha).
2025 police-recorded road crashes, spatially joined to 57 Prague municipal districts. The local metric is the share involving injury or death, stabilized with 20 crash equivalents at the city mean. It is not crime or individual crash probability. 11,029 crashes join exactly one municipal district, including 2,095 injury/fatal crashes. The police Prague jurisdiction has 11,014 crashes and is not identical to the municipal boundary; four jurisdiction-coded points do not join. No points are snapped to a district. Sparse districts are shrunk toward the city injury share; raw aggregate counts are retained. Reporting rules changed during 2025, including the property-damage reporting threshold. Compare only this same-period recorded-crash series. IPR municipal boundaries retrieved 2026-10-27. Topology-preserving simplification with area drift below 0.5%. Bruce explicitly approved limited aggregate use on 2026-10-27: publish self-calculated area aggregates and scores with official attribution; do not republish incident records, report prose or images. This is a project source-admission decision, not a claim of an open license. The citywide score uses Czechia national priors. Local evidence adjusts only mobility.
Natural=1.6; Current Conflict Intensity=1.5; Infrastructure=0.8; Access to health care=0.5. Published 0–10 components multiplied by ten for the existing research risk scale.; Country humanitarian-risk priors, not measured local street risks. European Union, CC BY 4.0.
Homicide: 0.8 per 100,000 (2023); theft: 325.8 per 100,000 (2022). Country observations.; Weak property calibration: existing France research risk 45 + bounded(10 × log2(325.8/2011.4), ±20), using the same-year UN theft series. Definitions and reporting differ between countries.
43 deaths per million residents, converted to 4.3 per 100,000. Preliminary estimate. European Union, CC BY 4.0.
2025 police-recorded road crashes, spatially joined to 57 Prague municipal districts. The local metric is the share involving injury or death, stabilized with 20 crash equivalents at the city mean. It is not crime or individual crash probability.; 11,029 crashes join exactly one municipal district, including 2,095 injury/fatal crashes. The police Prague jurisdiction has 11,014 crashes and is not identical to the municipal boundary; four jurisdiction-coded points do not join. No points are snapped to a district.; Sparse districts are shrunk toward the city injury share; raw aggregate counts are retained. Reporting rules changed during 2025, including the property-damage reporting threshold. Compare only this same-period recorded-crash series.; Bruce explicitly approved limited aggregate use on 2026-10-27: publish self-calculated area aggregates and scores with official attribution; do not republish incident records, report prose or images. This is a project source-admission decision, not a claim of an open license.