Identifying the bipolar (BP) spectrum, including the classic Bipolar I subtype (BP-I), Bipolar II (BP-II) and subthreshold bipolar disorders not meeting DSM-IV diagnostic criteria has raised growing interest, as these softer expressions of bipolar spectrum have been underdiagnosed in spite of clinical consequences. Data are from the Sao Paulo Epidemiological Catchment Area Study (N=1464). Non-affective controls were compared to BP spectrum groups, based on DSM-IIIR and on the "clinical significance" criteria: Subsyndromal Hypomania (SSH) and Manic Symptoms (MS). The lifetime prevalence of BP subgroups was 8.3% (N=122). All BP-I and -II and around 75% of SSH and MS subjects had a lifetime depressive syndrome. Compared to controls and MS subjects, BP-I, BP-II and SSH groups searched more medical help and mental health services. SSH group displayed higher rates of clinical significance than BP-I subjects, and suicidality was higher in BP groups compared to controls. Even the softer MS group had higher rate of suicide attempts than SSH subjects. This is a cross-sectional study and interviews were conducted by lay personnel. Replication in bigger community samples using a mood spectrum approach is necessary to confirm these findings. However, our findings were very similar to those obtained by other authors. Softer expressions of BP disorders appear in 6.6% of this community sample and have serious clinical consequences, which supports the importance of including these categories in the BP spectrum.