Objective: Previous estimates of the prevalence of seasonal affective disorder (SAD) in community-based samples generally originated from western countries. We report prevalence rates in eight groups from four latitudes in Turkey. Method: Seasonal Pattern Assessment Questionnaire (SPAQ) was distributed to the community-based samples from eight different locations at four latitudes in Turkey. The prevalence rates of winter SAD and subsyndromal SAD (S-SAD) were estimated for the four groups at the same latitudes by using SPAQ responses. Results: We distributed 3229 SPAQs, had an overall response rate of 54.16% and 1749 SPAQs were included in the analyses. Seasonality was reported as a problem by 549 subjects (31.57%) of our 1749 respondents. Prevalence of winter SAD and S-SAD are estimated as 4.86 and 8.35%, respectively, for the whole group. Prevalence rates were determined for each center and for four latitudes (two centers at the same latitude were grouped as one). In Adana-Gaziantep (It. 37), Izmir-Elazig (It. 38), Eskisehir-Ankara (it. 39) and Trabzon-Edirne (it. 41), the prevalence rates for winter SAD were 6.66, 2.25, 8.00 and 3.76%, respectively. Conclusions: Our prevalence estimates of winter SAD are similar to those found in previous community-based studies at the same latitudes; no correlation was found between latitude and prevalence of winter SAD, which could be related to the sampling methodology or to the fact that there were only 5degrees of difference between the latitudes. (C) 2002 Elsevier Science B.V. All rights reserved.