The analysis of natural disasters in a timely manner often suffers from limited sensor data. This limitation could be alleviated by leveraging information contained in images of the event posted on social media platforms, so-called “Volunteered Geographic Information (VGI)”. To save the analyst from manual inspection of all images posted online, we propose to use content-based image retrieval with the possibility of relevance feedback for retrieving only relevant images of the event. To evaluate this approach, we introduce a new dataset of 3,710 flood images, annotated by domain experts regarding their relevance with respect to three tasks (determining the flooded area, inundation depth, water pollution). We compare several image features and relevance feedback methods on that dataset, mixed with 97,085 distractor images, and are able to improve the precision among the top 100 results from 55% to 87% after 5 rounds of feedback.