Patterns of seasonal changes in mood and behavior in Montgomery County, Maryland, were evaluated in randomly selected household samples by lay interviewers using a telephone version of the Seasonal Pattern Assessment Questionnaire. The method for selecting the sample unit was random-digit dialing. We found that 92% of the survey subjects noticed seasonal changes of mood and behavior to varying degrees. For 27% of the sample seasonal changes were a problem and 4.3% to 10% of subjects, depending on the case-finding definition, rated a degree of seasonal impairment equivalent to that of patients with seasonal affective disorder. The seasonal pattern of "feeling worst" exhibited a bimodal distribution with a greater winter and a substantially lower summer peak (ratio, 4.5:1). Younger women who have a problem with seasonal changes and who feel worse on short days tended to exhibit the highest seasonality scores. It is apparent from our study that seasonal affective disorder represents the extreme end of the spectrum of seasonality that affects a large percentage of the general population. The influence of environmental factors on mood disorders and mood changes in the general population might provide valuable insight into pathogenesis, treatment, and prevention of affective illness.