In our previous work, acupuncture induced inhibition of the nociceptive respons e to peripheral noxious stimulation at spinal level could be blocked by iontopho retic bicuculline(Bic), an antagonist for GABA A receptors, suggesting an invol vement of GABA in acupuncture induced segmental inhibition. However some report s declare that increase in GABA content in brain is not benefit to acupuncture ana lgesia. In this paper, the effects of intra cerebroventricular and intrathecal i njection (icv and ith) of Bic on acupuncture analgesia were further investigated by tail flick latency tests in rats. There were 6～8 rats in each group. Th e results were as follows: 1. After icv GABA in dose of 0, 125, 250, 500 and 1000 μg /5 μL, the pain thre shold (PT) were raised to 102%, 108%, 128%, 136%, and 157% of the basal control value respectively. It indic ates that icv GABA could produce dose dependent analgesic effect. 2. After icv Bic at the dose of 10 and 20 μg /5 μL, icv GABA (500 μg /5 μL) induced analgesic effect lowered from 136.24±1.96% to 111.8±0.98% and 111.25±0.65% separately. It means that icv GABA induced analgesic effect is significantly blocked by preinjection of Bic. 3. After electro acupuncture at bilateral "Ciliao" (BL 32) points (50 Hz, 1～2 mA) for 10 minutes, PT were raised to 142.50±2.10% without any treatment and 1 4 3.72±2.04% with pretreatment of saline. When pretreated with icv Bic at doses o f 10 and 20 μg /5 μL, acupuncture analgesic effects (PT still raised to 141.74 ±1.54% and 146.71±1.85%) showed no significant reduction. It indicates tha t GABA in brain might not be involved in acupuncture analgesia. 4. After electro acupuncture at bilateral "Ciliao"(BL 32), PT were raised to 139.56±1.21% with ith saline and to 138.96±1.43% pretreated with ith Bic 5 μg /10 μL. When pretreated with ith Bic at doses of 10 and 20 μg /10 μL, PT wer e raised to 126.55±1.73% and 114.52±1.68% respectively, indicating the signifi cant reduction of acupuncture analgesia. It means that GABA mig
