Introduction In this study, we aimed to establish: (1) whether social isolation modifies the effect of unemployment on first episode psychosis and duration of untreated psychosis (DUP); and (2) whether the gap between high employment expectations and perceived poor employment achievement is associated with first-episode psychosis; and (3) whether the relationship of this achievement-expectation gap and first-episode psychosis is strongest in the African-Caribbean population. Method All patients with a first episode of psychosis presenting to specialist mental health services within tightly defined catchment areas in south-east London and Nottingham over a 2-year period were included in the study. A random sample of healthy participants living within the same catchment areas was also recruited. Data were collected on socio-demographic and clinical characteristics, DUP, social contacts, and perceived levels of employment achievement and expectation. Analysis was conducted on data of 546 participants (224 cases, 322 controls) from the SOP study. Results The relationship between unemployment and risk of non-affective psychosis was moderated by social contacts (unemployed/low social contacts, OR 7.52, 95% CI 2.97-19.08; unemployed/medium social contacts, OR 3.27, 95% CI 1.66-6.47; unemployed/high social contacts, OR 1.36, 95% CI 0.47-3.93). Unemployed patients experienced a longer DUP when having reported lower levels of social contacts. Participants whose employment achievement was lower than their expectations were more likely to be cases than those in whom achievement matched or exceeded expectations (adjusted OR 1.84, 95% CI 1.13-3.02). This applied equally to both African-Caribbean and White British participants (the Mantel-Haenszel test for homogeneity of odds ratios, chi(2) = 0.96, P = 0.33). Conclusions This study suggests that unemployment, social isolation, employment achievement and expectations are important environmental factors associated with risk of psychosis. More attention needs to be focused on interactions between environmental factors as well as subjective experience of those factors in future research on the aetiology of psychosis.