Classical commitment studies are either sociologically oriented or based in the psychological empirical research tradition. A review of the literature reveals agreement on the multidimensionality and the contextual complexity of commitment, two principles which guide the hypotheses of the present study. This study uses the Administrator and Teacher Survey (ATS) data from the High School and Beyond (HSB) project. These data permit measurement of teaching commitment in two dimensions: extra involvement in subject preparation and/or student activities; and strong identification with school staff educational goals, academic excellence, student personal growth and development, or with all three. Findings indicate that, despite its disadvantages, the ATS allows measures of commitment that may have some advantages over those found in either of the classiscal commitment research traditions. The present study allows for measurement of involvement and identification dimensions; it measures commitment as multidimensional; it allows the definition of commitment in a specific and appropriate social context; and it allows for possible conflicting underlying dimensions of commitment. The findings suggest that high schools present different social contexts and, therefore, policies on teaching commitment need to be approached on a local level. (AMH)