This study used a trait-state-occasion (TSO) model to isolate stable trait variance, occasion-specific state variance, and shared method related variance in a measure for leisure satisfaction in a Dutch nationally representative nine-year panel study. Findings indicate that satisfaction with leisure time is a consistently stronger indicator of overall leisure satisfaction than satisfaction with leisure activities. About half of the variance in leisure satisfaction is stable trait variance, with the remaining variance being mostly occasion-specific and to a lesser extent attributable to shared method variance and error. However, these findings depend on the age group we consider. Several socio-demographic variables relate directly to the trait aspect of leisure satisfaction. Our study underscores the importance of recognizing that over time leisure satisfaction measurements have considerable stable and more volatile elements and that one should control for shared method effects.