Two hundred and six patients, diagnosed with primary breast or prostate cancer completed self-report questionnaires on two occasions: before treatment (T1) and 12 months later (T2). The questionnaires included: the Hospital Anxiety and Depression Scale; Impact of Events Scale; the Metacognitions Questionnaire-30 (MCQ-30) and the Illness Perceptions Questionnaire-revised. A series of regression analyses indicated that metacognitive beliefs at T1 predicted between 14 and 19 % of the variance in symptoms of anxiety, depression and trauma at T2 after controlling for age and gender. For all three outcomes, the MCQ-30 subscale 'negative beliefs about worry' made the largest individual contribution with 'cognitive confidence' also contributing in each case. For anxiety, a third metacognitive variable, 'positive beliefs about worry' also predicted variance in T2 symptoms. In addition, hierarchical analyses indicated that metacognitive beliefs explained a small but significant amount of variance in T2 anxiety (2 %) and T2 depression (4 %) over and above that explained by demographic variables, T1 symptoms and T1 illness perceptions. The findings suggest that modifying metacognitive beliefs and processes has the potential to alleviate distress associated with cancer.