Work engagement, as conceptualized by the Utrecht Work Engagement Scale (UWES), is a work-related positive state of mind that is characterized by vigour, dedication and absorption; however, it still remains unclear how many dimensions work engagement comprises. This study tested the factorial validity of 11 different UWES factorial models that are available in the literature using a confirmatory factor analysis approach on a large sample of multi-occupational Polish employees (N = 1420). The two-factor UWES-6, comprising vigour and dedication, was found to be reliable and remained invariant across samples drawn from four different organizational positions. It also presented a better fit than the three-factor UWES-9, which is a scale used as a standard in contemporary research on work engagement. The findings suggest that the three-factor UWES-9 might not be an optimal measure of work engagement in Poland.