Background Residential mobility during upbringing, and especially adolescence, is associated with multiple negative mental health outcomes. However, whether associations are confounded by unmeasured familial factors, including genetic liability, is unclear. Aims We used a population-based case-cohort study to assess whether polygenic risk scores (PRSs) for schizophrenia, bipolar disorder and major depression were associated with mobility from ages 10-14 years, and whether PRS and parental history of mental disorder together explained associations between mobility and each disorder. Method Information on cases (n= 4207 schizophrenia,n= 1402 bipolar disorder,n= 18 215 major depression) and a random population sample (n= 17 582), born 1981-1997, was linked between Danish civil and psychiatric registries. Genome-wide data were obtained from the Danish Neonatal Screening Biobank and PRSs were calculated based on results of separate, large meta-analyses. Results PRSs for schizophrenia and major depression were weakly associated with moving once (odds ratio 1.07, 95% CI 1.00-1.16; and odds ratio 1.10, 95% CI 1.04-1.17, respectively), but not twice or three or more times. Mobility was positively associated with each disorder, with more moves associated with greater risk. Adjustment for PRS produced slight reductions in the magnitude of associations. Adjustment for PRS and parental history of mental disorder together reduced estimates by 5-11%. In fully adjusted models mobility was associated with all three disorders; hazard ratios ranged from 1.33 (95% CI 1.08-1.62; one move and bipolar disorder) to 3.05 (95% CI 1.92-4.86; three or more moves and bipolar disorder). Conclusions Associations of mobility with schizophrenia, bipolar disorder and depression do not appear to be attributable to genetic liability as measured here. Potential familial confounding of mobility associations may be predominantly environmental in nature.