To observe whether acupuncture up-regulates chemokine CXC ligand 1 (CXCL1) in the brain to play an analgesic role through CXCL1/chemokine CXC receptor 2 (CXCR2) signaling in adjuvant induced arthritis (AIA) rats, so as to reveal its neuro-immunological mechanism underlying improvement of AIA. BALB/c mice with relatively stable thermal pain reaction were subjected to planta injection of complete Freund adjuvant (CFA) for establishing AIA model, followed by dividing the AIA mice into simple AF750 (fluorochrome) and AF750+CXCL1 groups (n=2 in each group). AF750 labeled CXCL1 recombinant protein was then injected into the mouse's tail vein to induce elevation of CXCL1 level in blood for simulating the effect of acupuncture stimulation which has been demonstrated by our past study. In vivo small animal imaging technology was used to observe the AF750 and AF750+CXCL1-labelled target regions. After thermal pain screening, the Wistar rats with stable pain reaction were subjected to AIA modeling by injecting CFA into the rat's right planta, then were randomized into model and manual acupuncture groups (n=12 in each group). Other 12 rats that received planta injection of saline were used as the control group. Manual acupuncture (uniform reinforcing and reducing manipulations) was applied to bilateral "Zusanli" (ST36) for 4×2 min, with an interval of 5 min between every 2 min, once daily for 7 days. The thermal pain threshold was assessed by detecting the paw withdrawal latency (PWL) using a thermal pain detector. The contents of CXCL1 in the primary somatosensory cortex (S1), medial prefrontal cortex, nucleus accumbens, amygdala, periaqueductal gray and rostroventromedial medulla regions were assayed by using ELISA, and the expression levels of CXCL1, CXCR2 and mu-opioid receptor (MOR) mRNA in the S1 region were detected using real time-quantitative polymerase chain reaction. The immune-fluorescence positive cellular rate of CXCL1 and CXCR2 in S1 region was observed after immunofluorescence stain. The immunofluorescence double-stain of CXCR2 and astrocyte marker glial fibrillary acidic protein (GFAP) or neuron marker NeuN or MOR was used to determine whether there is a co-expression between them. In AIA mice, results of in vivo experiments showed no obvious enrichment signal of AF750 or AF750+CXCL1 in any organ of the body, while in vitro experiments showed that there was a stronger fluorescence signal of CXCL1 recombinant protein in the brain. In rats, compared with the control group, the PWL from day 0 to day 7 was significantly decreased (P<0.01) and the expression of CXCR2 mRNA in the S1 region significantly increased in the model group (P<0.05), while in comparison with the model group, the PWL from day 2 to day 7, CXCL1 content, CXCR2 mRNA expression and CXCR2 content, and MOR mRNA expression in the S1 region were significantly increased in the manual acupuncture group (P<0.05, P<0.01). Immunofluorescence stain showed that CXCR2 co-stained with NeuN and MOR in the S1 region, indicating that CXCR2 exists in neurons and MOR-positive neurons but not in GFAP positive astrocytes. Acupuncture can increase the content of CXCL1 in S1 region, up-regulate CXCR2 on neurons in the S1 region and improve MOR expression in S1 region of AIA rats, which may contribute to its effect in alleviating inflammatory pain.