The best case series (BCS) is a retrospective chart review that describes a series of patients who all appear to have benefitted from the treatment under study. The BCS has been advocated as the first research step for evaluating complementary and alternative medicine (CAM) treatments for cancer. However, the research value of the BCS has not been assessed. To address this deficiency, the present study evaluates the primary characteristics of the BCS process through a systematic review of the English language scientific literature. Twenty-four individual BCS investigating 16 unique CAM treatments for cancer were identified. About half of the BCS reported evidence of tumor regression in association with a particular CAM treatment, but only six contained documentation adequate for publication in peer-reviewed journals. For these six BCS the number of responders per BCS ranged from 2 to 12 (median, 3.5), the proportion of responders in the total number of evaluated cases varied from 6% to 100% (median, 40%), and the proportion of evaluated cases to identified cases ranged from 18% to 53% (median, 29%). The primary factors confounding the identified BCS were lack of documentation of disease and/or the use of concurrent or recent conventional treatment. Despite these general deficiencies, four BCS (antineoplastons, hydrazine sulfate, laetrile, and Kelly-Gonzalez) were sufficiently convincing to warrant follow-up clinical trials. These data suggest that while well-documented BCS do have an impact on the research agenda, in general, additional rigor is needed during their compilation.