Ten men, with or without alopecia, were observed for a period of between 8 and 14 years using phototrichograms on a precisely located zone on the vertex of the scalp. Among the various parameters observed, we chose the percentage of hairs in telogen as the criterion for assessment of hair shedding. Mathematical analysis of the variations in this telogen percentage was carried out for each individual subject and for the whole group, as represented by the population mean (or the 'average subject'). This analysis demonstrated the existence of overall annual periodicity, manifested by a maximal proportion of telogen hairs at the end of summer and the beginning of autumn. Some subjects also exhibited a periodicity approximately corresponding to two annual peaks. In those subjects with a very low proportion of hairs in telogen, no periodicity was demonstrated. In another group of subjects, it has been shown that the variations in telogen percentage reflect those observed in hair shedding, assessed in a standardized manner. Periodicity of the telogen percentage, and hence of hair fall, is not independent of climatic factors (sunshine hours), and these must be taken into account when assessing the treatment or prevention of hair loss.