This study examines the relationship between dimensions of job and non‐job activity, job satisfaction and mental health among veterinary professionals and tests specific predictions arising from the work of Broad bent (1985) and Karasek (1979). Results are based on an analysis of 411 questionnaires returned in a postal survey. Job and non‐job activities contributed a significant proportion of unique variance in total mental health as measured, and related differently to anxiety and depression. Support for Broadbent's (1985) predictions was mixed. Pacing related distinctively to anxiety, but depression showed no relationship to social aspects of activity. Both lack of control over the speed of activities and discretion were related to mental health indices, but support was found for the role of discretion as a buffer in an interactive sense on one outcome measure only. Results are discussed in terms of ideas arising from the clinical literature, and suggestions offered for the potential use of carefully chosen activity in controlling levels of anxiety, depression and well‐being.