To explore the mechanism of electroacupuncture (EA)-induced cumulative analgesic effects on chronic pain in rats with or without ovariectomy (OVX). A total of 110 female Wistar rats were randomized into normal control (n=10), chronic constrictive injury (CCI, n=10), CCI+EA (n=30), OVX+CCI (n=30), and OVX+CCI+EA (n=30) groups. Each of the latter 3 groups was further divided into 2 days (2 d), 2 weeks (2 W) and 3 weeks (3 W) subgroups, respectively (n=10 in each subgroup). The CCI pain model was established by ligature of the right sciatic nerve, and the memory impairment model duplicated by OVX. The paw withdrawal latency (PWL, pain threshold) of the bilateral footplates was detected by radiant heat irradiation, and the bilateral difference in PWL (PWLD) was used to evaluate changes in the pain reaction. Morris water maze test was conducted for evaluating the rats' learning-memory ability. EA was applied to bilateral Zusanli (ST36) and Yanglingquan (GB34) for 2 d, 2 W and 3 W, respectively. Pituitary and hypothalamic beta-endorphin (EP) and adrenocorticotrophic hormone (ACTH) contents were detected by immunoradioassay. Compared with the CCI group, PWLD of the CCI+EA-3 W group decreased significantly (P<0.05). Compared with the OVX+CCI group, PWLD of the OVX+CCI+EA-3 W group was lowered considerably (P<0.05), but the value was markedly higher than its basal value and those of the normal control and CCI+EA groups (P<0.05). In comparison with the sham-OVX group, the escape latency, swimming distance (SD) in the target quadrant and total SD were increased remarkably in the OVX group (P<0.05), while the number of target platform crossings was decreased significantly (P<0.05), suggesting an impairment of the OVX rats' learning-memory ability. In simple CCI rats, both beta-EP and ACTH contents of the pituitary increased markedly (P<0.05), and those of the hypothalamus decreased obviously compared to the normal control group (P<0.05). After EA, pituitary and hypothalamic ACTH levels were significantly lowered at 2 d and hypothalamic ACTH and beta-EP contents increased obviously at 3 W in comparison with the CCI group (P<0.05). In OVX+CCI rats, following EA, pituitary beta-EP contents at 2 d, 2 W and 3 W, and hypothalamic beta-EP and ACTH contents at 2 W and hypothalamic ACTH levels at 3 W increased significantly (P<0.05), but hypothalamic beta-EP level at 3W decreased markedly (P<0.05). The effects of repeated EA in lowering pituitary ACTH and raising hypothalamic beta-EP and ACTH levels disappeared after OVX+CCI. Repeated EA has a cumulative analgesic effect, which is closely associated with its effects in regulating pituitary and hypothalamic beta-EP and ACTH levels. OVX may weaken the analgesic effect of EA by affecting hypothalamic-pituitary axis activity.