Thirty-six subjects affected with hypersomnia associated with mood disorders, 31 with a diagnosis of dysthymia, 4 with a diagnosis of bipolar disorder and one with a diagnosis of major recurrent depression underwent standardized polysomnographic procedures including night 1, MSLT and night 2 (uninterrupted). 36.1% of these subjects had a reduced or intermediate mean sleep latency on the MSLT and 13.8% slept over 9 hr at night. In addition 17 of these subjects underwent prolonged polysomnography during day 2. In comparison with eight subjects affected with idiopathic hypersomnia, mean sleep latency on the MSLT was significantly longer and total sleep time during night 2 and during night 2 plus day 2 was significantly lower in subjects affected with hypersomnia associated with mood disorders. It is concluded that a positive diagnosis of hypersomnia associated with a mood disorder requires both behavioral observation and polysomnography. Among these subjects there may be subjects with well-documented hypersomnia and subjects with anergia facilitating or mimicking sleep.