Objectives. Past studies expected measures of obesity to be positively associated with positive affects. However, this hypothesis was not tested in reference to a specific positive affect. We tested the hypothesized unidirectional effects of measures of obesity on vigour, representing a positive affect, and of vigour on measures of obesity. Design. We used a longitudinal design, separately for men and women. Participants were 1,876 and 931 healthy men and women, respectively, examined at Time 1 (T1) and Time 2 (T2), about 2 years apart. Methods. Measures of obesity included body mass index, waist circumference, and waist-to-hip ratio. Vigour was assessed by the Shirom-Melamed Vigour Measure. We used structural equation modelling to test our hypotheses. Results. We found that for both genders, T1 measures of obesity did not predict either T1 or T2 vigour. Among both genders, we found support for the effects of T1 vigour on T1 but not on T2 measures of obesity. Conclusions. To the extent that the 'Jolly fat hypothesis' refers to the effects of measures of obesity on positive affects, we failed to support it for vigour as a positive affect. Vigour has contemporaneous but not longitudinal effects on body weight.