Source-linked overall safety research estimate with an area safety layer; lower confidence reflects geographic fallbacks and incomplete local dimensions.
Safrom's citywide research estimate is 81/100. Higher scores mean safer on the model's scale. Evidence as of . Confidence: C.
What does this area's map cover?
20 areas with estimates. Area evidence as of .
injury_crashes · 2024-01-01 / 2024-12-31.
City violence and property inputs use ten separate SSMSI source indicators, bounded relative to the existing Paris research anchor. Only injury-crash pressure varies within this area map. Other dimensions inherit the city estimate. Resident population is not visitor exposure; road counts have no traffic-flow denominator. No street, hotel or night safety is established.
City violence and property inputs use ten separate SSMSI source indicators, bounded relative to the existing Paris research anchor. Only injury-crash pressure varies within this area map. Other dimensions inherit the city estimate. Resident population is not visitor exposure; road counts have no traffic-flow denominator. No street, hotel or night safety is established.
Locate your intended stay on the map and check whether that area has evidence.
Read what drives the local differences. A citywide score and an area estimate answer different questions.
Check current local advice and your transport options separately. These estimates cannot assess a specific hotel, street or time of day.
Citywide evidence
Personal violence
83/100
Property & scams
54/100
Getting around
88/100
Health & rescue
91/100
Conflict & stability
86/100
Disaster & environment
85/100
Infrastructure response
91/100
City violence and property inputs use ten separate SSMSI source indicators, bounded relative to the existing Paris research anchor. Only injury-crash pressure varies within this area map. Other dimensions inherit the city estimate. Resident population is not visitor exposure; road counts have no traffic-flow denominator. No street, hotel or night safety is established.
Ten source indicators use separate like-for-like rate ratios against Paris; counts with different units are never added.; Resident-normalised records are not traveller victimisation probabilities; unreported crime and different visitor volumes remain limitations.; Local curve is a bounded research hypothesis, not an officially published safety score.
does not isolate Paris; U.S.-traveler perspective; advisory level is not an absolute safety score; narrative guidance requires structured extraction review
country-level only; modeled perceptions rather than traveler incidents; cannot represent neighborhood variation or current events; confidence intervals must not be discarded
BAAC counts police-recorded road crashes with at least one person requiring care; property-damage-only crashes are outside this dataset.; One characteristics row per unique accident ID; not a victim, vehicle or all-incident count. Records missing coordinates, outside the boundaries or on an ambiguous boundary are excluded, never moved to the nearest area.; Absolute counts are affected by roads, traffic and pedestrian/cyclist exposure. These are mobility evidence, not crime rates or a safety ranking.; 2024 facts are shown using the current 20-quartier boundary file; the source catalogue historical description of six sectors is not used as the feature count.