OBJECTIVE To investigate seasonal patterns in mood and behavior and estimate the prevalence of seasonal affective disorder (SAD) and subsyndromal seasonal affective disorder (S-SAD) in the Icelandic population. METHODS A random sample generated from the Icelandic National Register, consisting of 1000 men and women aged 17 to 67 years from all parts of Iceland. It represents 6.4 per million of the Icelandic population in this age group. METHODS The Seasonal Pattern Assessment Questionnaire, an instrument for investigating mood and behavioral changes with the seasons, was mailed to a random sample of the Icelandic population. The data were compared with results obtained with similar methods in populations in the United States. METHODS Seasonality score and prevalence rates of seasonal affective disorder and subsyndromal seasonal affective disorder. RESULTS The prevalence of SAD and S-SAD were estimated at 3.8% and 7.5%, respectively, which is significantly lower than prevalence rates obtained with the same method on the east coast of the United States (chi 2 = 9.29 and 7.3; P < .01). The standardized rate ratios for Iceland compared with the United States were 0.49 and 0.63 for SAD and S-SAD, respectively. No case of summer SAD was found. CONCLUSIONS Seasonal affective disorder and S-SAD are more common in younger individuals and among women. The weight gained by patients during the winter does not seem to result in chronic obesity. The prevalence of SAD and S-SAD was lower in Iceland than on the East Coast of the United States, in spite of Iceland's more northern latitude. These results are unexpected since the prevalence of these disorders has been found to increase in more northern latitudes. The Icelandic population has remained remarkably isolated during the past 1000 years. It is conceivable that persons with a predisposition to SAD have been at a disadvantage and that there may have been a population selection toward increased tolerance of winter darkness.