The Bamako Initiative, a controversial attempt to strengthen Primary Health Care using community financing and community participation and management was launched at a meeting of African Ministers of Health in 1987. This evaluation focuses particularly on the community financing aspects of the Initiative. Previous experiences of community financing highlight particular lessons for the development of the Initiative and issues likely to be encountered in implementation attempts. Four country case studies of the Initiative at its early stages of implementation (in June, July and August 1991) were conducted in Burundi, Guinea, Kenya, and Nigeria. Similar activities in Uganda were also studied. These were short studies of one month each and utilised methods of rapid evaluation. The studies aimed to direct implementers quickly to issues requiring attention rather than to reach overall conclusions regarding the success of the Initiative which would be premature at this stage. Price structures used by the Initiative need to consider the access of marginalised groups more than is the case at present. In addition, there is a need to ensure against over-prescription and commercialisation and to ensure that incentives for utilising most appropriate levels of care are maintained. The evidence suggests that most people do find amounts of money to pay for health services which are large in relation to their income. This is probably a tribute to extensive community support mechanisms. Nevertheless, it highlights the plight of those who fall through this safety net for whom even charges for very basic care may be prohibitive. On the other hand, it appears that in most cases, the Initiative's activities provide a service which is cheaper when all costs to the household are taken into account, than was available before. The quality of services included in the Initiative's activities in the five countries was highly variable. Success in raising substantial revenues has also been mixed. In some countries however, substantial funds have been generated and used to achieve real improvements in health services. Overall, the experiences of the five countries appear to have been highly dependent on a number of 'environmental' characteristics: a tradition or not of 'free' services; the adequacy of current resource availability and that immediately preceding the introduction of the Initiative; the existing stage and nature of decentralisation within the country; and the competition the Initiative's activities face with alternatives.