Associated with increasing urbanization worlwide, artificial light at night is increasingly recognized as a threat to wild animals. In birds, light pollution has been shown to affect behaviour and physiology. However, most field studies so far have been correlational and short-term, thus it is unclear if light pollution can ultimately affect fitness and population size. Here we present a comprehensive study on the long-term effects of light pollution on the great tit (Parus major), a songbird common in European cities. We used an experimental approach in the field where lamp posts with LED of different colors (white, green and red) were installed in previously dark forests in the Netherlands. Using a 8-year dataset of breeding events, mark-recaptures, behavioural and physiological data, we first show that exposure to experimental light pollution in the field result to changes in energy expenditure, nocturnal activity, spatial movements, susceptibility to infection and availability of caterpillar preys. Most of these effects were color-dependent, with white and green light having stronger effects than red light. However, behavioural and physiological changes were not followed by fitness effects, as reproductive output and survival probability were not affected by any of the light types. This suggests that either behavioural and physiological effects were not stronger enough to lead to fitness costs, or that they were in fact adaptive responses to artificial light. Alternatively, light pollution could lead to non-random settlement patterns, so that only birds with specific traits are able to breed under bright night lights. Further research should focus on distinguishing between these different options to fully elucidate the long-term implications of light pollution for wild populations.