The predictability of turnover was compared for part-time versus full-time workers. This was accomplished using variables (1) that were hypothesized by Mobley (1977) to precede the turnover decision and (2) that have been shown in previous research to be significantly related to turnover. Measures of these variables were assessed via questionnaires following the second month of employment. Turnover was assessed directly from company records for a period of up to twelve months following employment. None of the hypothesized antecedent variables were significantly related to turnover within the part-time group. However, within the full-time group, all of the hypothesized precursors to turnover were found to be at least marginally related to actual turnover. Further, there was evidence to suggest differential predictability of turnover across full-time and part-time employment status groups, suggesting that the two groups may have a different ‘psychology of work’. These results are discussed in terms of the supportive evidence provided.