A priori cannabis use was recorded at index admission for 112 participants in the Manchester first-episode psychosis cohort. 69 of the 100 surviving (mainly schizophrenia) patients were followed up 10-12 years later and assessed on a battery of clinical, behavioural and neurocognitive measures. Individuals who had not used cannabis before the first episode of illness were generally indistinguishable from cannabis users at follow-up, except that the latter group evidenced a marked 'sparing' of neurocognitive functions. These findings are briefly discussed in relation to other casual factors in psychosis.