The links between unionisation and job satisfaction remain controversial. In keeping with the existing literature we find strong statistically significant negative correlations between unionisation and overall job satisfaction. However, in contrast to the previous literature we find that once one accounts for fixed unobservable differences between covered and uncovered employees, union coverage is positively and significantly associated with satisfaction with pay and hours of work. Failure to account for fixed unobservable differences between covered and uncovered employees leads to a systematic underestimate of the positive effects of coverage on job satisfaction for both union members and non-members. It seems union coverage has a positive impact on job satisfaction that is plausibly causal.