Research studies have found an association between occupational stress and psychosomatic symptoms, psychological problems, and behavioral maladjustment. Very few studies on job stress have focused on school administrators. This research was designed to explore the impact of occupational stress on public school administrators in Minnesota. Significant differences between administrative levels were hypothesized for the following variables: twenty individual occupational stressors, an index of global occupational stress, length of work week, and four measures of facet-free job satisfaction;This cross-sectional study of school administrator stress employed a self-administered questionnaire consisting of seventy-seven items. The scales and individual questions from earlier research on occupational stress, job satisfaction, and related variables were field tested and refined through two pilot projects;Four hundred fourteen of the 424 superintendents employed by public schools in Minnesota were mailed the questionnaire. A computer generated random selection procedure was employed to select 225 secondary school principals and 225 elementary school principals from the membership lists of the two principal organizations. Seven hundred forty-eight administrators responded with a valid questionnaire for a return rate of 87 percent. The data were analyzed using descriptive statistics, chi square analysis, and Pearson product moment correlations;Statistically significant differences (p < 0.01) between the three samples of administrators were found on eight of the twenty occupational stressors. On a global measure of occupational stress, no significant differences were obtained. Forty to 49 percent of each sample reported their work environment was either "always" or "usually" stressful. There were no significant differences between the three stratified samples on four measures of facet-free job satisfaction. Seventy-three percent of all administrators were satisfied with their administrative position, 11 percent were neither satisfied nor dissatisfied, and 16 percent were dissatisfied;A correlation matrix analysis identified the following clusters of stress variables: role conflict; responsibility for people; work overload and role expansion; special education and legal compliance; and conflicting demands. The stress cluster of responsibility for people had the highest ranking individual stressors of any of the clusters for both samples of principals. For superintendent respondents, the highest stress was experienced by the cluster of conflicting demands;In conclusion, on most of the independent variables, there were no significant differences between the three administrative samples. There was a slight tendency for secondary principals to report more occupational stress than either elementary principals or superintendents. In a concluding section, a variety of stress reduction programs were recommended.