Depression is a polygenic trait that causes extensive periods of disability and increases the risk of suicide, a leading cause of death in young people. Previous genetic studies have identified a number of common risk variants which have increased in number in line with increasing sample sizes. We conducted a genome-wide association study (GWAS) in the largest single population-based cohort to date, UK Biobank. This allowed us to estimate the effects of ≈ 20 million genetic variants in 320,000 people for three depression phenotypes: broad depression, probable major depressive disorder (MDD), and International Classification of Diseases (ICD, version 9 or 10) coded MDD. Each phenotype was found to be significantly genetically correlated with the results from a previous independent study of clinically defined MDD. We identified 12 independent loci that were significantly associated (P < 5 × 10 -8 ) with broad depression, two independent variants for probable MDD, and one independent variant for ICD-coded MDD. Gene-based analysis of our GWAS results with MAGMA revealed 52 regions significantly (P < 2.72 × 10 -6 ) associated with broad depression, six regions in probable MDD and three regions in ICD-coded MDD. Gene region-based analysis of our GWAS results with MAGMA revealed 57 regions significantly (P < 6.01 × 10 -6 ) associated with broad depression, of which 35 were also detected by gene-based analysis. Variants for broad depression were enriched in pathways for excitatory neurotransmission and neuron spines. This study provides a number of novel genetic risk variants that can be leveraged to elucidate the mechanisms of MDD and low mood.