Rats were classified into either acupuncture effective animals or non-effective animals by a significant increase (p < 0.05) in tail-flick latency. This classification is proper for 0.5 mg/kg morphine analgesia and dorsal periaqueduct central gray stimulation-produced analgesia (d-PAG-SPA). Individual variation in effectiveness of acupuncture and morphine analgesia was correlated, each having the value r=0.75. Effective animals contain higher concentrations of endogenous morphine-like factor (MLF) than non-effective animals. MLF unit activity was 25 times higher in effective animals than non-effective animals. Close correlation having the value r=0.78 was observed between individual MLF contents and individual effectiveness of acupuncture analgesia. The plain pain threshold was not correlated with the contents of MLF. Concentrations of opiate receptor were not correlated with individual variations in effectiveness of acupuncture and morphine analgesia. A dose of 250 mg/kg i.p. D-phenylalanine treatment enhanced analgesia induced by intraventricular applied met5-enkephalin and beta-endorphin. Acupuncture, morphine analgesia and d-PAG-SPA were enhanced to a similar degree in both acupuncture effective and non-effective animals after D-phenylalanine. Hence the individual variation in effectiveness of analgesia was abolished by treatment of D-phenylalanine.