Background and Aims: The use of cannabis has previously been linked to both depression and self-harm; however, the role of genetics in this relationship is unclear. This study aimed to estimate the phenotypic and genetic associations between cannabis use and depression and self-harm. Design: Cross-sectional data collected through UK Biobank were used to test the phenotypic association between cannabis use, depression and self-harm. UK Biobank genetic data were then combined with consortia genome-wide association study summary statistics to further test the genetic relationships between these traits using LD score regression, polygenic risk scoring and Mendelian randomization methods. Setting: United Kingdom, with additional international consortia data. Participants: A total of 126 291 British adults aged between 40 and 70 years, recruited into UK Biobank. Measurements: Phenotypic outcomes were life-time history of cannabis use (including initial and continued cannabis use), depression (including single-episode and recurrent depression) and self-harm. Genome-wide genetic data were used and assessment centre, batch and the first six principal components were included as key covariates when handling genetic data. Findings: In UK Biobank, cannabis use is associated with an increased likelihood of depression [odds ratio (OR) = 1.64, 95% confidence interval (CI) = 1.59–1.70] and self-harm (OR = 2.85, 95% CI = 2.69–3.01). The strength of this phenotypic association is stronger when more severe trait definitions of cannabis use and depression are considered. Using consortia genome-wide summary statistics, significant genetic correlations are seen between cannabis use and depression [rg = 0.289, standard error (SE) = 0.036]. Polygenic risk scores for cannabis use and depression explain a small but significant proportion of variance in cannabis use, depression and self-harm within a UK Biobank target sample. However, two-sample Mendelian randomization analyses were not significant. Conclusions: Cannabis use appeared to be both phenotypically and genetically associated with depression and self-harm. Limitations in statistical power mean that conclusions could not be made on the direction of causality between these traits.