The key focus of this longitudinal study in the Netherlands was to determine the role of social support (i.e. perceived availability of emotional support, lack of received problem-focused emotional support, and negative interactions) and positive and negative self-esteem in depressive symptoms in 475 recently diagnosed cancer patients and 255 individuals without cancer from the general population. Patients and the comparison group were interviewed and filled in a questionnaire at two points in time: 3 months (T1) and 15 months (T2) after diagnosis. The results indicated that social support and self-esteem were weakly to moderately related to each other. Negative self-esteem was more strongly related to all three types of social support, compared to positive self-esteem. Regression analyses showed that social support and self-esteem were independently related to depressive symptoms (concurrently), such that lower levels of social support and self-esteem were strongly associated with higher levels of depressive symptoms. This finding suggests that these two resources supplement each other additively. A longitudinal analysis showed that social support and self-esteem also predicted future levels of depressive symptoms, although the explained variance was much lower than in a cross-sectional analysis. Comparisons between cancer patients and the comparison group generally revealed no significant differences between the two groups in the associations of social support and self-esteem with depressive symptoms. The only exception was a lack of problem-focused emotional support. At three months after diagnosis, a lack of this type of support, characterised by reassuring, comforting, problem-solving, and advice, was more strongly related to depressive symptoms in patients than in the comparison group. (C) 2003 Elsevier Science Ltd. All rights reserved.