This study tested the predicted relation of satisfaction to person-environment fit using an alternative approach (Edwards, 1991) to that typically used in theories of career choice and adjustment. Fuzzy work preferences, fuzzy job perceptions, and nonfuzzy job satisfaction data were collected for 21 work value and interest dimensions from up to 352 respondents in five different organizations. Hierarchical regression analyses were undertaken at the item level, examining the independent and interacting contribution of work preferences and job perceptions to satisfaction. Job perceptions accounted for most of the variance in job satisfaction while interactions with work preference (fit) added significant incremental variance for 13 of the 21 satisfaction items. The judged match index (Hesketh, McLachlan, & Gardner, 1992) explained further incremental variance for 3 of the items. The data demonstrate the importance of examining direct relations of person and situational variables as well as "fit" and the advantages in retaining information at the item level. Theoretical and practical implications of the results are discussed.