Low levels of some polyunsaturated fatty acids (PUFAs) could influence behaviors leading to the abuse of substances through their actions on serotonergic and dopaminergic mechanisms. Because substance abusers tend to have poor dietary habits, the possibility that a deficient intake of n-3 PUFAs, available from dietary sources only, and subsequent low n-3 plasma levels would predict their relapse rates was explored. Thirty-five patients admitted to substance abuse clinics were enrolled and followed for one year. Dietary questionnaires and blood samples were collected at baseline and on a quarterly basis, and relapse rates monitored on a monthly basis. Six patients dropped out shortly after study entry, 11 relapsed in the course of the study and dropped out, 7 relapsed but completed the study, and 11 did not relapse and completed the study. Non-relapsers were found to have significantly higher levels of docosahexaenoic acid (DHA) calculated as microg/ml and % TFA, when compared to relapsers (p = .031 and p = .010, respectively) and to relapsers and non-completers combined (p = .014 and p = .009, respectively). These pilot data suggest, but do not prove, the existence of a relationship between low levels of DHA and relapse vulnerability in some individuals who abuse substances. The study of the efficacy of n-3 supplements or of dietary modifications on relapse appears warranted.