Background: Depression has profound effects on an individual's life but evidence relating to the causal effects on health is limited. Previous studies examining the relationship between depression and disease are largely observational making causal prediction challenging due to confounding and reverse causation. The aim of this study is to establish genetic evidence for a causal association between depression and a range of disease outcomes. Methods We performed hypothesis-free phenome-wide association analyses of depression and 925 disease outcomes based on linked data on hospitalisations and mortality from 337,536 participants in the UK Biobank. Depression was instrumented by weighted genetic risk score (GRS) derived from 44 depression variants, with sensitivity analyses using restricted 17 variant score. GRSs-disease outcome associations passing the multiple-testing corrected significance threshold (P<1.9 x 10-3) were followed by Mendelian randomization (MR) analyses. Findings: Depression GRS was associated with 42 disease outcomes in the phenome-wide discovery stage, and inverse-variance weighted (IVW) MR analyses supported a possible causal link between depression and 25 of these disease outcomes. The strongest signal was observed for depression diagnosis (Pdiscovery<8·4 x 10-14), and related co-morbidities including anxiety (Pdiscovery= 3·0 x 10-7; MR IVW OR 1·89, 95%CI 1·50 to 2·38), and sleep disorders (Pdiscovery=1·8 x 10-4; OR 1·59, 95%CI 1·27 to 1·98). Further genetic evidence for a possible causal effect of depression was seen for inflammatory and haemorrhagic gastrointestinal diseases, including diseases of the oesophagus (Pdiscovery=3·0 x10-7; OR 1·30, 95%CI 1·17 to 1·46), non-infectious gastroenteritis (Pdiscovery=1·2x10-3; OR 1·25, 95%CI 1·06 to 1·48), and gastrointestinal haemorrhage (Pdiscovery=3·8x10-4; OR 1·26, 95%CI 1·11 to 1·43). Signals were also observed for symptoms/disorders of the urinary system (Pdiscovery=1·7 x 10-5; OR 1·36, 95%CI 1·19 to 1·56), asthma (Pdiscovery =3·9 x 10-4; OR 1·23, 95%CI 1·06 to 1·44), and painful respiration (Pdiscovery =4·0 x10-7; OR1·28, 95%CI 1·14 to 1·44). Depression was also causally linked with disorders of lipid metabolism (Pdiscovery=3·8x10-5; OR 1·22, 95%CI 1·12 to 1·34) and ischaemic heart disease (Pdiscovery=3·3x10-7, OR 1·30, 95CI 1·15 to 1·47). Seven variants showed evidence for pleiotropy, but sensitivity analyses excluding these variants provided consistent associations. Interpretation Our study indicated a causal link between depression and a broad range of diseases, suggesting a notable burden of co-morbidity. Early detection and management of depression is important, but treatment strategies should be selected to also minimize the risk of related co-morbidities. Funding: Australian Research Training Program. Conflict of interest: The authors have declared that no competing interest exist. Ethical Approval Statement: We have used an open-access database, called UK biobank, to generate our finding and conclusion. UK Biobank has approval from the North West Multi-centre Research Ethics committee, Patient Information Advisory Group, and Community Health Index Advisory Group (http://www.ukbiobank.ac.uk/ethics/). Each participant provided written consent during the recruitment phase of the project (http://www.ukbiobank.ac.uk/wp-content/uploads/2011/06/Consent_form.pdf?phpMyAdmin=trmKQlYdjjnQIgJ%2CfAzikMhEnx6).