For an occupation to become a profession, it is often contended that its members must possess ‘vocational commitment’. Yet, there has in fact been little research either in New Zealand or overseas on the whole question of vocational or professional commitment. The present research was primarily concerned with the type and extent of commitment in those undergoing training for a profession. It consisted of a longitudinal study of a cohort of student nurses and student teachers who entered training in a provincial New Zealand city in 1973. A scale to measure expressed vocational commitment (EVC) at the time of entry was devised and was applied to the complete sample of trainee teachers and nurses on the first day of their training. Additional data were gathered early in the training course via standardised personality, psychological and attitude inventories, as well as background details on such variables as sex, socio-economic status and ethnicity. The research then focussed on the direction of, and stated reasons for, changes in EVC during the course of training. To facilitate this, a subsample was selected from the two groups for regular interviewing and observation. As well, students leaving the course were interviewed, and where ‘hunches’ were formed from either subsample or leaver material, an endeavour to check these was made through discussion with a monitoring group and by gathering additional evidence from other members of the role set. At the end of their training period the surviving sample was post-tested with an adapted EVC scale. The research demonstrated the complexity of abstract concepts such as vocational commitment, and confirmed that it is multidimensional in nature. Moreover, and perhaps most significantly, the findings showed that EVC was subject to considerable fluctuations over relatively short periods of time, which suggested that commitment itself has substantial situationally-specific components. The research also demonstrated that most students on entry to training were extremely tentative in their career choice, and in some instances were merely trying out possibilities. In addition many had poorly formed perceptions of what the vocation they were to enter would require of them, both as students and as practitioners. Although the analysis endeavoured to pinpoint reasons for changes in EVC, this task proved extraordinarily difficult. It brought out some of the shortcomings of the positivistic style of research, and much of the theorising associated with it. From a more practical viewpoint the research established the difficulties associated with predicting eventual success in a training programme, even when a wide range of entry variables is available for analysis. Some important implications for teacher and nurse education can be adduced, particularly the value of deferring the final selection of applicants during training, thus drawing on an older more mature pool. The research findings also indicate that improved orientation programmes prior to entry to formal training are needed; that a diversified and flexible training course is required to cater for the wide range of the students’ abilities, aptitudes and attitudes; and that improved guidance and counselling services for some ‘at risk’ students are highly desirable. Overall, the conclusion was drawn that vocational commitment in the two groups studied was idiosyncratic in its development and influence. Although exceptions could be found, students were more reflective about their decision making and in their career planning than many previous researchers had suggested.