Several studies in the Amazonian Basin have shown that riverine populations are exposed to methylmercury through fish consumption. It has been suggested that seasonal variations in hair mercury observed through sequential analyses may be related to the changes in fish species ingested by the local communities. The aim of the present study was to investigate the relationship between fish-eating practices and seasonal variation in mercury exposure. A group of 36 women from a village located on the banks of the Tapajos River, a major tributary of the Amazon, comprised the present study population. An interview-administered questionnaire was used to gather information on socio-demographic characteristics, fish-eating practices and other relevant information. The women also provided hair samples of at least 24 cm in length for mercury analysis. Hair total and inorganic mercury concentration was measured using a cold vapor atomic absorption analytical method. Trigonometric regression analysis was done to assess the seasonal variation of total mercury levels. Variations in inorganic mercury were examined by repeated measures analysis of variance, and analysis of contrast variable with a polynomial transformation. The results showed that hair mercury levels varied with the season. Higher levels were observed in months corresponding to the dry season, with lower levels in the rainy season. Herbivorous fish predominated the diet for 47.2% of the women during the dry season, but this rose to 72.2% during the rainy season. Those who reported eating fish daily had higher mercury levels in hair compared to those who only ate fish a few times per week. Retrospective mercury analyses, evaluated by the quantity of mercury present in each centimeter of hair, indicate that mean mercury level of the population decreased over the 2 years prior to the study. The percentage of inorganic mercury over the total mercury in hair increased towards the extremities of the hair strand. Higher percentages of inorganic mercury were found for the group who ate more fish (on a daily consumption basis). These results support the assumption that there are seasonal variations in methylmercury exposure and also a relationship between type of fish species consumed and the resulting hair mercury levels.