In 2025, Berlin police recorded 25,811 bike thefts worth €32.8 million in declared value. This is where, when, and how the city loses its bikes, drawn live from the raw police open-data records.


In 2025, Berlin police logged 25,811 bike thefts, one roughly every 20 minutes, all year round. Just 137 were failed attempts; the rest succeeded.
€32.8Min declared value stolen in one year, an average of €1,281 per bike.
Python · D3 · Berlin Polizei open data (2025)
At planning-area resolution the pattern is unmistakable: theft concentrates in the dense inner city and the eastern Kieze, thinning toward the outer edges. This is the exact choropleth from the original QGIS analysis, shaded by thefts per Planungsraum.
Drill down to planning-area level and a few names dominate. Alt-Treptow tops the list, a bike stolen roughly every 1.9 days, followed by Oranienburger Straße and Humboldthain. High-footfall corridors around stations and nightlife generate a strikingly outsized share of all thefts.
Zoom back out and the concentration is stark. Just three of Berlin's twelve districts, Mitte (17%), Friedrichshain-Kreuzberg (15%) and Pankow (12%), account for 44% of every bike theft in the city.
Read the heatmap and two patterns jump out. The top rows, Mitte, Friedrichshain-Kreuzberg, Pankow, glow brightest. And the darkest column sits at 18:00: theft peaks in the early evening, not at night. Bikes are taken while owners run their last errand of the day.
Switch the ranking between raw theft count and total declared value, the order shifts as wealthier districts lose fewer but pricier bikes.
The most-stolen categories are ordinary men's and women's city bikes, the workhorses of daily Berlin life. High-value road and cargo bikes are rarer targets but pull up the average loss.
Declared values span sub-€100 commuters to €5,000+ cargo and sporting models. The median loss is €900; the mean €1,281, pulled up by a long tail of expensive e-bikes and cargo bikes.
€900median declared value per stolen bike.
Data & method. 25,811 individual theft records from the Berlin Police open-data portal (Fahrraddiebstahl, calendar year 2025), each with location (LOR planning unit), timestamp, bike type and declared value. Aggregated in Python; charts rendered live in D3. Districts derived from the two-digit LOR prefix. A granular Planungsraum-level choropleth is in progress.
← Back to all data stories