The ecological impact of artificial light at night (ALAN) is an increasingly recognized process that accompanies expanding urbanization. Yet, we have limited knowledge on the impact of ALAN on wild species, and on the potential to mitigate any negative effects by using different light sources and colors. In birds, effects of ALAN on activity levels are reported for several species and, hence, their daily energy expenditure (DEE) may be affected. DEE is a potent mediator of life-history trade-offs and fitness and thus an important aspect to consider when examining the potential long-term ecological effects of ALAN. Previous work has suggested that birds exposed to ALAN show higher levels of provisioning and nocturnal activity, suggesting that white ALAN increases DEE. Other factors regulating DEE, such as provisioning behavior and food availability, might also respond to ALAN and thus indirectly affect DEE. We tested the hypothesis that ALAN increases DEE using an experimental setup where four previously unlit transects were illuminated with either white, green, or red LED light, or left dark as a control treatment. This setup was replicated in eight locations across the Netherlands. We measured DEE of our focal species, the great tit (Parus major), using a novel doubly labeled water technique that uses breath rather than blood samples. Contrary to our expectations, birds feeding their offspring under white and green ALAN showed lower DEE compared to birds in the control dark treatment. Differences in chick provisioning activity did not explain this result, as neither visit rates nor daily activity timing was affected by light treatment. However, food availability under white and green light was much higher compared to red light and the dark control. This difference strongly suggests that the lower DEE under white and green ALAN sites is a consequence of higher food availability in these treatments. This result shows that there can be positive, indirect effects of ALAN for breeding song birds which may balance against the negative direct effects shown in previous studies.