Presents findings from the Saxony Longitudinal Study pertaining to personality correlates and gender invariance of item wording effects using the German version of the Rosenberg Self-Esteem Scale (RSES), which is designed to assess subjective self-worth. Annually collected data were available for 346 adults (aged 38-39) who had completed their baseline assessment in 1987 at the age of 14-15 years. Participants also completed the Big Five Inventory - short version. Since prior evaluations of RSES dimensionality have not consistently yielded single-factor solutions, it is suggested that item wording may confound its factorial structure. Latent variable analyses yielded a two-factor solution with one substantive and two specific factors related to positive and negative wording. This model performed better than nine other alternative ones and partial scalar invariance was observed for gender. Despite insignificant gender differences in average RSES scores, latent mean comparisons did indicate a slight male advantage. Implications for the use of personality factors as predictors of latent factors are discussed.