Using two single-cohort longitudinal surveys, the NLS72 and the NELS88, I investigate the impact of four noncognitive traits--self-esteem, external locus of control, the importance of money/work and the importance of people/family--on wages and on the gender wage gap among these young workers. I find that gender differences in these noncognitive factors, especially the importance of money/work, have a modest but significant role in accounting for the gender wage gap. Methodologically, this paper proposes a correction to the Oaxaca-Blinder-Ransom decomposition that results in a truly decomposable approach compatible with the simple pooled regression that includes a gender dummy. (Contains 7 tables and 41 footnotes.)