We previously reported the anti-hyperalgesia of electroacupuncture (EA) on persistent inflammatory pain in an unrestrained, unsedated, and conscious rat model. Using this model, induced by injecting complete Freund's adjuvant (CFA) into one hind paw, we systematically evaluated the anti-hyperalgesia of EA stimulation parameters (frequency, intensity, treatment duration, and pulse width). We assessed hyperalgesia by paw withdrawal latency (PWL) to a noxious thermal stimulus and found that 10- and 100-Hz EA frequencies at a current intensity of 3 mA produced the greatest anti-hyperalgesia, when compared to other parameters. Both frequencies significantly increased PWL in the early phases of hyperalgesia (2.5 and 24 h; p < 0.05), and 10 Hz EA also significantly increased PWL in later phases (5 to 7 days; p < 0.05). A sufficient but tolerable intensity of 3 mA was more effective than lower intensities (1-2 mA). A 20-min treatment produced better anti-hyperalgesia than longer and shorter (10 and 30 min) treatments. Acupoint specificity study demonstrated that GB30 produced significant EA anti-hyperalgesia, while Waiguan (TE5) and sham points, an abdominal point and a point at the opposite aspect of GB30, did not. The spinal Fos protein expression study demonstrated that the optimal EA selectively suppressed Fos expression in superficial laminae (I/II) and activated it in deeper laminae (III/IV) of the spinal dorsal horn. The results suggest that the EA anti-hyperalgesia is parameter-dependent and point-specific, and they provide important information for designing further clinical acupuncture research on persistent inflammatory pain.