Chloroquine resistance was first detected in Kenya in 1978 and escalated during the 1980s. Chloroquine remained the treatment of choice for uncomplicated malaria infections until revised guidelines were launched in 1998 despite a plethora of scientific evidence on failure. This review analyses the range and quality of the evidence base that was used to change the drug policy in Kenya from chloroquine to SP and examines the process of consensus building and decision making. Our review illustrates the difficulties in translating sensitivity data with gross geographical, temporal and methodological variations into national treatment policy. The process was complicated by limited options, unknown adverse effects of replacement therapies, cost, as well as limited guidance on factors pertinent to changing the drug policy for malaria. Although > 50% of the studies showed parasitological failures by 1995, there was a general lack of consensus on the principles for assessing drug failures, the inclusion criteria for the study subjects and the relative benefits of parasitological and clinical assessments. A change in international recommendations for assessment of drug efficacy in 1996 from parasitological to clinical response further perplexed the decisions. There is an urgent need for international standards and evidence-based guidelines to provide a framework to assist the process by which decision-makers in malaria-endemic countries can make rational choices for antimalarial drug policy change.