Background Manual acupuncture (MA) with different stimulus frequencies may give rise to varying acupuncture effects. However, the intensity-effect relationship and the underlying mechanisms of MA remain unclear. Objective To compare the analgesic effects of different frequencies of twirling-rotating MA on rats with complete Freund's adjuvant- (CFA-) induced pain and explore the underlying mechanism via peripheral sensory nerves. Methods First, 36 healthy male Wistar rats were randomly divided into 6 groups: control group, 2 r/s MA group (twirling-rotating MA with the frequency of 2 revolutions per second), 4 r/s MA group (twirling-rotating MA with the frequency of 4 revolutions per second), CFA group, CFA + 2 r/s MA group, and CFA + 4 r/s MA group. Rats in three CFA groups received an intraplantar injection of CFA to establish a pain model, while the rats in other three groups received an intraplantar injection of saline. Rats in the 2 r/s MA group and 4 r/s MA group were treated with the corresponding frequencies of twirling-rotating MA on bilateral Zusanli (ST36) and Kunlun (BL60) for 7 days. The ipsilateral nociceptive thresholds (paw withdrawal latency; PWL) were tested to evaluate the analgesic effects. Second, 9 healthy male Wistar rats were randomly divided into 3 groups: control group, 2 r/s MA group, and 4 r/s MA group. The proportion of C-fiber neurons (calcitonin gene-related peptide- (CGRP-) positive neurons) and A-fiber neurons (neurofilament 200- (NF200-) positive neurons) in the dorsal root ganglia (DRG) activated by MA were quantitatively analyzed with the morphological immunofluorescence staining method. Third, 30 healthy male Wistar rats were randomly divided into 6 groups: control group, CFA group, CFA + 2 r/s MA group, CFA + 2 r/s MA + RTX group, CFA + 4 r/s MA group, and CFA + 4 r/s MA + RTX group. Resiniferatoxin (RTX) was injected into the acupoints before acupuncture. PWL was evaluated to investigate the analgesic effect. Results Both types of MA treatment increased the PWL of saline-injecting rats and pain model rats. Moreover, 4 r/s MA was superior to 2 r/s MA in increasing PWL. A higher quantity of excited C-fiber neurons was observed following 4 r/s MA than 2 r/s MA, while the reverse was observed in the activation of A-fiber neurons. Following the injection of RTX to inhibit the activation of C-fibers, the analgesic effect of 4 r/s MA reduced significantly but not of 2 r/s MA. Conclusion Strong MA (4 r/s MA) has superior analgesic effects to gentle MA (2 r/s MA) on CFA model rats, which is associated with C-fiber activation.