Sleep disturbance affects about 75% of depressed individuals and is associated with poorer patient outcomes. The genetics in this field is an emerging area of research. Thus far, only core circadian genes have been examined in this context. We expanded on this by performing a genome-wide association study (GWAS) followed by a preplanned hypothesis-driven analysis with 27 genes associated with the biology of sleep. All participants were diagnosed by their referring physician, completed the Beck Depression Inventory (BDI), and the Udvalg for Kliniske Undersogelser Side Effect Rating Scale at baseline. Our phenotype consisted of replies to 3 questions from these questionnaires. From standard GWAS chip data, imputations were performed. Baseline total BDI scores (<i>n</i> = 364) differed significantly between those with and those without sleep problems. We were unable to find any significant GWAS hits although our top hit was for changes in sleep and an intergenic marker near <i>SNX18</i> (<i>p</i> = 1.06 × 10^–6). None of the markers in our hypothesis-driven analysis remained significant after applying Bonferroni corrections. Our top finding among these genes was for rs13019460 of Neuronal PAS Domain Protein 2 with changes in sleep (<i>p</i> = 0.0009). Overall, both analyses were unable to detect any significant associations in our modest sample though we did find some interesting preliminary associations worth further exploration.