The excess mortality in people with psychotic disorders is a major public health concern, but little is known about the clinical and social risk factors which may predict this health inequality and help inform preventative strategies. We aimed to investigate mortality in a large epidemiologically characterized cohort of individuals with first-episode psychosis compared with the general population and to determine clinical and social risk factors for premature death. All 557 individuals with first-episode psychosis initially identified in 2 areas (Southeast London and Nottinghamshire, United Kingdom) were traced over a 10-year period in the ӔSOP-10 study. Compared with the general population, all-cause (standardized mortality ratio [SMR] 3.6, 95% confidence interval [CI] 2.6-4.9), natural-cause (SMR 1.7, 95% CI 1.0-2.7) and unnatural-cause (SMR 13.3, 95% CI 8.7-20.4) mortality was very high. Illicit drug use was associated with an increased risk of all-cause mortality (adj. rate ratio [RR] 2.31, 95% CI 1.06-5.03). Risk of natural-cause mortality increased with a longer time to first remission (adj. RR 6.61, 95% CI 1.33-32.77). Family involvement at first contact strongly reduced risk of unnatural-cause mortality (adj. RR 0.09, 95% CI 0.01-0.69). Our findings suggest that the mortality gap in people with psychotic disorders remains huge and may be wider for unnatural-cause mortality than previously reported. Efforts should now focus on further understanding and targeting these tractable clinical and social risk factors of excess mortality. Early intervention and dual diagnosis services may play a key role in achieving more rapid remission and carer involvement and addressing substance use problems to reduce excess mortality in psychosis.