Background: Etiologic differences between bipolar I disorder (BD-I) and bipolar II disorder (BD-II) have been challenged recently, and family epidemiologic studies may elucidate the matter. Nevertheless, it remains unclear whether BD-I and BD-II display different familial aggregation patterns within each bipolar disorder subtype and coaggregation with other psychiatric disorders.Method: Per the Taiwan National Health Insurance Research Database (N = 23,258,175), patients with bipolar disorder were classified as having BD-I or BD-II based on the history of psychiatric hospitalization for a manic episode. During the study period (2001-2011), 184,958 first-degree relatives (FDRs) of patients with BD-I and BD-II were identified. By comparing patients with 1:4 age-, sex-, and kinship-matched samples without BD-I/BD-II probands, the relative risks (RRs) of major psychiatric disorders were estimated.Results: FDRs of BD-I probands had a significantly higher risk of BD-I than those of BD-II probands (BD-I proband: RR = 15.80 vs BD-II proband: RR = 5.68, P < .001). The risk of BD-II was similar between FDRs of BD-I and BD-II probands (BD-I proband: RR = 6.48 vs BD-II proband: RR = 5.89, P = .1161). Familial aggregation was greater within each BD subtype than among cross-subtypes. Furthermore, FDRs of BD-I probands had an increased risk of schizophrenia (BD-I probands: RR = 5.83 vs BD-II probands: RR = 2.72, P < .001); FDRs of BD-II probands had a higher likelihood of attention-deficit/hyperactivity disorder (BD-II probands: 2.36 vs BD-I probands: 1.93, P = .0009).Conclusions: The risk of psychiatric disorders is higher among the FDRs of patients with either BD-I or BD-II. Furthermore, the familial specificity of BD-I and BD-II assessed in this study may further the current understanding of etiologic boundaries between bipolar disorder subtypes.