We were fortunate in having access to a survey of 800 adults in Nanjing that was conducted by the International Wool Secretariate in 1986. The survey was intended for market research (e.g., brand recognition), but questions were asked about the respondent's age, occupation, income, and education level. These questions, though not ideally worded, were sufficient for the undertaking of a returns-to-education study. We believe the data to be unique and, of course, the issue of interest is whether a member of the workforce in China faces investment decisions on education similar to those faced elsewhere. The influence of sex and marital status on the level of income achieved by an individual was also examined. The basic equation modeled personal income as a nonlinear function of schooling and experience, but multicollinearity problems were encountered. As a result, a production function approach was successfully adapted to the returns-toeducation model. The collinearity problem was overcome, and this approach allowed the ease of substitution of education for experience in income generation to be examined more formally. An additional problem of errors in variables arose from a failing in questionnaire design and was handled statistically.