Background. Individuals with autism spectrum disorder (ASD) appear to be at increased risk of non-affective psychotic disorder (NAPD) and bipolar disorder (BD). However, most previous studies examined the co-occurrence of ASD and NAPD or BD, ignoring possible diagnostic bias and selection bias. We used longitudinal data from Dutch psychiatric case registers to assess the risk of NAPD or BD among individuals with ASD, and compared the results to those obtained for the Dutch population in earlier studies. Methods. Individuals with ASD (n = 17 234) were followed up between 16 and 35 years of age. Kaplan-Meier estimates were used to calculate the risk of NAPD or BD. We conducted separate analyses to reduce possible bias, including an analysis among individuals diagnosed with ASD before age 16 years (n = 8337). Results. Of the individuals with ASD, 23.50% (95% confidence interval 21.87-25.22) were diagnosed with NAPD and 3.79% (3.06-4.69) with BD before age 35 years. The corresponding figures for the general population were 0.91% (0.63-1.28) and 0.13% (0.08-0.20). Risk estimates were substantially lower, but still higher than general population estimates, when we restricted our analyses to individuals diagnosed with ASD before age 16, with 1.87% (1.33-2.61) being diagnosed with NAPD and 0.57% (0.21-1.53) with BD before age 25 years. The corresponding figures for the general population were 0.63% (044-0.86) and 0.08% (0.05-0.12). Conclusions. Individuals with ASD are at increased risk of NAPD or BD. This is likely not the result of diagnostic or selection bias.