INTRODUCTION: Treatment adherence plays a pivotal role in hospitalisation in bipolar disorder (BD). We examined the impact of adherence and pharmacological variables on involuntary vs. voluntary admission on a sample of inpatients from the European Mania in Bipolar Longitudinal Evaluation of Medication study (EMBLEM). METHODS: 1374 inpatients with an acute manic or mixed episode of BD participated in this observational study on clinical, functional and economic outcomes of pharmacological treatment. We analysed data at the time of study inclusion, and the primary outcome measure was admission status (voluntary vs. involuntary admission). RESULTS: The strongest baseline factor of admission status was adherence whereby patients' adherence was significantly associated with admission when treated with atypical antipsychotics or lithium as monotherapy. Adherence with typical antipsychotics was not significantly associated with admission status. DISCUSSION: These results emphasise the crucial role of treatment adherence for admission status and, within the context of a naturalistic study, some advantage of atypical over typical antipsychotics on admission in acute mania of BD.