ObjectiveTo determine the factors that influence condom use among prostitutes and their clients in The Gambia.DesignA cohort of 181 prostitutes working in seven bars and several rural markets in The Gambia were monitored daily for 14 months. A sample of 747 clients of these prostitutes was also questioned.Main outcome measureProportion of sexual contacts for which a condom was used.ResultsData on 24 181 sexual contacts reported by the prostitutes indicated condom use varied according to type of partner (from 84% with clients to only 4% with regular partners). Condom use with clients varied according to location (from 91% in high-class bars to 59% in rural markets), decreased from 91% with the first client of the evening to 37% with the tenth client, and from 75% with clients paying higher charges (> D19) to 52% with those paying lower charges (< D20). Condom use was not related to the socio-demographic characteristics of the prostitutes. Clients reported lower condom use than prostitutes. Clients aged 20–24 years were least likely to use condoms, while white collar workers, traders, and those paying higher charges, were more likely to use condoms.ConclusionsThe level of condom use in this cohort of prostitutes was high but not consistent. Condom use was determined more by the type of establishment and the characteristics of clients, than by any fixed tendency among the prostitutes. Thus, education campaigns should be directed as much to clients as to prostitutes. In The Gambia, ‘lower-class' bars and those in rural areas where prostitutes work should be a priority target.