Prior research has documented linear detrimental effects of commuting on individuals’ life satisfaction: the longer individuals’ daily commute, the less satisfied they are with their life. An inspection of the available longitudinal evidence suggests that this conclusion is almost exclusively based on a continuous operationalization of commuting time and distance with a focus on a linear relationship. In contrast, cross-sectional evidence indicates preliminary evidence for non-linear effects and suggests that negative effects of commuting are particularly likely when commuting exceeds a certain threshold of time or distance. Relying on nationally representative data for Germany, the present study applies longitudinal modelling comparing estimates from a continuous and a categorical operationalization. Results clearly indicate a non-linear association and show that negative effects of commuting are almost completely due to individuals who commute more than 80 km (50 miles) daily per way. These findings are in conflict with prior research (partly resting on the same data) proposing a linear relationship. Further analyses suggest that satisfaction with leisure time is a significant mediator of the observed non-linear effect. Results are discussed in light of prior theorizing on the consequences of commuting.