Objectives Dispositional optimism is often considered to be a unidimensional construct. Recent studies suggest, however, that optimism and pessimism are separate dimensions. In this study we investigated two issues. First, the levels of optimism and pessimism in adolescents with cancer compared with healthy controls and second, the individual effects of optimism and pessimism on concurrent and longitudinal well-being. Design A matched case-control design was used to examine whether adolescents with cancer and healthy adolescents differed with regard to optimism and pessimism. The second part of the study was employed in a prospective design with assessments in the patient group at 3 and 6 months post-diagnosis. Methods Thirty-three adolescents with cancer (3 months post-diagnosis) and 66 matched controls completed a measure on dispositional optimism (i.e., optimism and pessimism). In addition, patients completed measures on positive and negative aspects of well-being at 3 and 6 months post-diagnosis. Results Although adolescents with cancer were not more optimistic than their healthy peers, they were significantly less pessimistic. Zero order and semi-partial correlations showed that optimism and pessimism are related to different aspects of well-being. Specifically, we found a cohesive pattern in which optimism predicts positive aspects and pessimism negative aspects of well-being. Conclusions The high levels of overall optimism often found in patients with cancer might in fact result from low pessimism instead of high optimism. Furthermore, as our study shows that optimism and pessimism are differentially associated with aspects of well-being, it provides strong support for the bidimensionality of dispositional optimism.