The association between presumed protective factors and social risk factors for hospitalization and mortality was studied during a 14-year follow-up period in a cohort of 8,168 Swedish men aged 18-20 years at baseline. Using Cox regression analysis, the authors found that five protective factors (high social class, home well-being, school well-being, good emotional control, and self-perceived good health) were associated with lower risks of hospitalization and death. Four social risk factors (contact with police or child welfare authorities, running away from home, having divorced parents, and ever using narcotics) were significantly associated with increased risk of hospitalization and mortality. The relative hazard decreased with the number of protective factors and increased with the number of social risk factors, almost linearly. The relative hazard was 0.24 for hospitalization among those with six protective factors and 0.24 for mortality for those with five or six protective factors. The relative hazard for hospitalization was 3.09 among those with five social risk factors compared with those with none, while for mortality the relative hazard among those with four or five social risk factors was 5.74 compared with those with none. While these results indicate strong cumulative effects for both the social risk factors and the protective factors, the associations of individual factors with the two outcome measures were generally reduced in models which simultaneously adjusted for all factors, which presumably indicates collinearity among the factors. There was only limited support for a buffering, or interacting, effect between the risk factors and the protective factors.