This article reports a longitudinal study of relative weight, smoking, and mental health as predictors of medically certified sickness and unauthorized absence from work among student nurses (N = 185). Information about smoking, relative weight, and self-reports of somatic complaints and social dysfunction was obtained prior to the 33-month period over which sickness and absence were recorded. Multiple regression was used to test a predictive model relating absence to linear and quadratic components of relative weight, smoking, and symptom measures. A significant curvilinear relation between relative weight and absence was found, the form of which closely resembled the relation between relative weight and mortality; smoking showed an additive effect. A linear interaction between social dysfunction and relative weight was also found; particularly high levels of absence occurred among those of high relative weight who also reported high levels of social dysfunction. Analysis of sickness episodes confirmed the adverse effects of overweight and, to a lesser extent, of underweight and smoking. The findings are discussed in terms of medical, psychological, and psychosocial influences on sickness and absenteeism.