Objectives. To evaluate the long-term (psychosomatic) health consequences of man-made earthquakes compared with a non-exposure control group. Exposure was hypothesised to have an increasingly negative impact on health outcomes over time. Setting. Large-scale gas extraction in the Netherlands causing earthquakes and considerable damage. Participants. A representative sample of inhabitants randomly selected from municipal population records; contacted 5 times during 21 months (T1: N=3934; T5: N=2150; mean age: 56.54; 50% men; at T5, N=846 (39.3%) had no, 459 (21.3%) once and 736 (34.2%) repeated damages). Main measures. (Psychosomatic) health outcomes: self-rated health and Mental Health Inventory (both: validated; Short Form Health Survey); stress related health symptoms (shortened version of previously validated symptoms list). Independent variable: exposure to the consequences of earthquakes assessed via physical (peak ground acceleration) and personal exposure (damage to housing: none, once, repeated). Results. Exposure to induced earthquakes has negative health consequences especially for those whose homes were damaged repeatedly. Compared with a no-damage control group, repeated damage was associated with lower self-rated health (OR:1.64), mental health (OR:1.83) and more stress-related health symptoms (OR:2.52). Effects increased over time: in terms of relative risk, by T5, those whose homes had repeated damage were respectively 1.60 and 2.11 times more likely to report poor health and negative mental health and 2.84 times more at risk of elevated stress related health symptoms. Results for physical exposure were comparable. Conclusion. This is the first study to provide evidence that induced earthquakes can have negative health consequences for inhabitants over time. It identifies the subpopulation particularly at risk: people with repeated damages who have experienced many earthquakes. Findings can have important implications for the prevention of negative health consequences of induced earthquakes.