This report presents information regarding the patterns of variation in the salaries paid to public and private school teachers in relation to various personal and job characteristics. Specifically, the analysis examines the relationship between compensation and variables such as public/private schools, gender, race/ethnic background, school level and type, teacher qualifications, and different work environments. The economic conceptual framework of hedonic wage theory, which illuminates the trade-offs between monetary rewards and the various sets of characteristics of employees and jobs, was used to analyze The Schools and Staffing Survey (SASS) database. The national survey was administered by the National Center for Education Statistics during the 1987-88, 1990-91, and 1993-94 school years. Findings indicate that on average, public school teachers earned between about 25 to 119 percent higher salaries than did private school teachers, depending on the private subsector. Between about 2 and 50 percent of the public-private difference could be accounted for by differences in teacher characteristics, depending on the private subsector. White and Hispanic male public school teachers earned higher salaries than their female counterparts. Hedonic wage theory would predict that teacher salaries would be higher in schools with more challenging, more difficult, and less desirable work environments. Schools with higher levels of student violence, lower levels of administrative support, and large class sizes paid higher salaries to compensate teachers for the additional burdens. However, some of the findings contradict the hypothesis. For example, public school teachers working in schools characterized by fewer family problems, higher levels of teacher influence on policy, and higher job satisfaction also received higher salaries. In conclusion, the results are consistent with the hypothesis that a complex array of factors underlie the processes of teacher supply and demand and hence the determination of salaries. Teachers are not all the same, but are differentiated by their attributes. At the same time, districts and schools are differentiated by virtue of the work environment they offer. Seventeen tables and two figures are included. Appendices contain technical notes, descriptive statistics and parameter estimates for variables, and standard errors for selected tables. (Contains 84 references.) (LMI)