Attrition analyses are reported from a longitudinal study which was concerned with the psychological impact of unemployment and unsatisfactory employment in a sample of Australian school leavers. In order to assess the internal validity of the study the following two questions were posed: (a) Does the attrition rate vary with employment status?, and (b) Are pretest scores for dropouts related to employment status? We found that the dropout rate was grater in the employed and dissatisfied employed than in the satisfied employed and full‐time tertiary Students. we conclude that the internal. validity of longitudinal studies investigating the psychological impact of unemployment or unsatisfactory employment in young people is partially compromised by selective attrition in that the most disaffected (the unemployed and the dissatisfied employed) are the most likely to drop out. On the other hand, these status‐related differences in attrition did not appear to be related either to demographic characteristics or to pretest scores on the dependent variables.