Aims: To determine whether the risk of psychotic symptoms during weeks of methamphetamine use was dependent on, increased by, or independent of having a family history of psychosis. Design: Secondary analysis of 13 contiguous 1-week periods of data (1370 weeks). A risk modification framework was used to test each scenario. Setting: Geelong, Wollongong and Melbourne, Australia. Participants: Participants in a randomized controlled trial of treatment for methamphetamine dependence (n = 148) who did not have a primary psychotic disorder on enrolment. Measurements: Psychotic symptoms in the previous week were defined as a score of 3+ on any of the Brief Psychiatric Rating Scale items of hallucinations, unusual thought content or suspiciousness. Any (vs no) methamphetamine use in the previous week was assessed using the Timeline Followback method. Self-reported family history of psychosis was assessed using the Diagnostic Interview for Psychosis. Findings: The risk of psychotic symptoms in the past week was independently associated with methamphetamine use in that week (relative risk [RR] = 2.3, 95% CI = 1.3–4.3) and with having a family history of psychosis (RR = 2.4, 95% CI = 0.9–7.0); the joint risk among participants with a family history of psychosis during weeks when they were using methamphetamine was large (RR = 4.0, 95% CI = 2.0–7.9). There was no significant interaction between a family history of psychosis and methamphetamine use in predicting psychotic symptoms (interaction RR = 0.7 95% CI = 0.3–1.8), but there was a small non-significant excess risk due to the interaction (0.20 95% CI = −1.63 to 2.03). Conclusions: Among people dependent on methamphetamine, the relative risk of psychotic symptoms during weeks of methamphetamine use does not appear to be dependent on, or increased by, having a family history of psychosis. However, a family history of psychosis does appear to be an independent risk factor that contributes to the absolute risk of psychotic symptoms in this population.