One-hundred and twelve cadets attending the 141st training course for reserve officers of the 'Scuola Transmission' of the Italian Army were administered, 10 days after incorporation, a battery of personality tests and measures of stressful events in the preceding year. Test scores were considered both individually and grouped into factors. During the 6 months of the course all disease episodes for each cadet were recorded. Total episodes infections and traumas were considered. A significantly higher number of total episodes and, specifically, of infections was present in subjects reporting a higher number of stressful events, in interaction with attitudes towards parent figures, hardiness, loneliness, and an alienation factor. A younger age of subjects also appeared predictive of a higher number of total episodes and infections, as a main effect and in interaction with attitudes to mother, hardiness, and alienation. Very few effects were obtained for traumas, suggesting that the effects of variables are mediated through a biological route rather than through behaviour, as would be the case if traumas were involved. No effect was shown by measures of stressful events alone. Results are discussed in the light of a reconsideration of the notion of stressful events.