Objective: Urban birth is associated with risk of non-affective psychoses. While emerging evidence suggests that this association is also apparent for subclinical positive psychotic phenomena in the general population, few studies have considered which specific aspects of the urban environment predict risk, or whether these factors also increase the likelihood of negative symptoms. Method: Using data from the Avon Longitudinal Study of Parents and Children, linked to census geographical indicators, we examined whether population density, deprivation, inequality, and social fragmentation at birth were associated with negative and positive psychotic symptoms at age 16 and 18 years, respectively. We used logistic regression models, controlling for child’s ethnicity, maternal age, education, marital status, social class, depressive symptoms, other neighbourhood exposures, and, in sensitivity analyses, polygenic risk scores (PRS) for schizophrenia in a sub-sample of children of white ethnicity (N=10,283). Results: Amongst 11,879 adolescents, those born in the most densely populated tertile had greater odds of reporting positive psychotic experiences, after multivariable adjustment (odds ratio (OR): 1.57, 95% confidence intervals (CI): 1.14 – 2.17). Adolescents born in the most socially fragmented neighbourhoods had greater odds of negative symptoms, after multivariable adjustment (OR: 1.43, 95%CI: 1.06 – 1.85). These associations were not confounded by schizophrenia PRS. There was no evidence of any other associations. Discussion: Birth into more densely populated and socially fragmented environments increased risk of positive and negative psychotic phenomena in adolescence. Our findings suggest that different forms of neighbourhood social adversity impinge on different psychopathophysiologies associated with the clinical expression of psychosis.