Background: Insomnia and hypersomnia are common in bipolar disorder (BD) but it is unclear whether genetic factors influence this association.Stratifying by bipolar subtypes could elucidate the nature of this association and inform research on sleep and BD. We therefore aimed to determine whether polygenic risk scores (PRS) for insomnia, daytime sleepiness and sleep duration differ according to bipolar subtypes (bipolar I disorder, BD-I or bipolar II disorder, BD-II). Methods: In this case-control study, we used multinomial regression to determine whether PRS for insomnia, daytime sleepiness, and sleep duration were associated with risk of BD-I or BD-II when compared to each other and to controls. Cases (n=4672) were recruited within the United Kingdom from the Bipolar Disorder Research Network. Controls (n=5714) were recruited from the 1958 British Birth Cohort and UK Blood Service. All participants were of European ancestry. Clinical subtypes of BD were determined by semi-structured psychiatric interview (the Schedules for Clinical Assessment in Neuropsychiatry) and case notes.Results: Within cases, 3404 and 1268 met DSM-IV criteria for BD-I and BD-II, respectively. Insomnia PRS was associated with increased risk of BD-II (RR = 1.14, 95% CI = 1.07-1.21, P = 8.26 x 10 -5 ) but not BD-I (RR = 0.98, 95% CI = 0.94-1.03, P = .409) relative to controls. In contrast, sleep duration PRS was associated with increased relative risk of BD-I (RR = 1.10, 95% CI = 1.06-1.15, P = 1.13 x 10 -5 ), but not BD-II (RR = 0.99, 95% CI = 0.93-1.06, P = .818). Daytime sleepiness PRS was associated with an increased risk of BD-I (RR = 1.07, 95% CI = 1.02-1.11, P = .005) and BD-II (RR = 1.14, 95% CI = 1.07-1.22, P = 3.22 x 10 -5 ) compared to controls, but did not distinguish