Most empirical analyses of earnings measure work experience as years since school graduation and cannot, therefore, separate age and experience effects. This study demonstrates that the return to work experience is highly significant in the presence of age variables, but that the returns to experience and schooling are sensitive to the way experience is measured. Initial employment experiences are shown to have a significant effect on subsequent earnings; failure to account for this in an experience measure attributes too much human capital development to nonwork activities and reduces our ability to explain the earnings of young men aged 16-32. (Author)