This paper reviews the various experimental and survey strategies employed to assess the relative importance of different job characteristics in determining job satisfaction, and analyzes the problems involved with the different approaches. The paper then describes the development of an empirically derived explanatory model of job satisfaction, taking into account problems of multi-co-linearity and interaction among the predictor variables. The importance of any job facet was equated with its capacity to account for variance in job worker satisfaction scores. The job facets were subsumable into four basic dimensions: (1) opportunities it provided to perform challenging or self-developing activities; (2) the resources enabling adequate performance; (3) provision of a comfortable work environment; and (4) provision of financial rewards and job security. The model was able to explain some 53% of the variance in job satisfaction scores, and the authors note some of the problems that make such explanatory power acceptable. Of the four dimensions, "challenge" seemed most promising. Further directions for research and refinement are indicated. (NG)