Rheumatoid arthritis is a symmetrical disease where inflammation spreads from one joint to involve the mirror image joint, a phenomenon also seen in animal models of arthritis (Donaldson et al. 1993). Clinical evidence from patients (Veale et al. 1993), and from experimental studies (Donaldson et al. 1995; Kidd et al. 1995), suggests that a neural mechanism contributes to the spread of arthritis. We investigated whether there are changes in antidromic neuronal activity in the homologous nerve contralateral to that innervating an inflamed joint, which could account for the spread of arthritis through a neurogenic mechanism. Complete Freund’s Adjuvant (CFA; 250μg/100μl) was injected intra-articularly into the right knee joint of male Wistar rats (n=20) under halothane anaesthesia (3% in O₂). Intra-articular injection of CFA resulted in significant swelling of the injected joint (p<0.001, vs contralateral, paired t test). 3-5 days post-CFA, rats (n=16) were re-anaesthetised (60mg/kg pentobarbital i.p.) and the external jugular vein and trachea cannulated. Extracellular recordings were made from teased filaments of the left saphenous nerve, contralateral to the inflamed knee joint, sectioned distally. Data are means ± SEM. The frequency of centrally generated spontaneous efferent activity recorded from the contralateral saphenous nerve was significantly higher in rats with CFA-induced inflammation (2.9 ± 1Hz, n=16) than that seen in control non-injected rats (0.1 ± 0.05Hz, n=5; p<0.01, Mann Whitney). Capsaicin (2%) applied directly onto the nerve proximal to the recording site in CFA rats (n=6) significantly reduced activity levels (0.2 ± 0.1Hz, 1-way ANOVA) indicating that the recorded activity was predominantly in VR1 expressing sensory neurons. In a separate set of experiments, CFA-inflamed rats (4 days) and control rats (n=4 each group) received an injection of Evan’s Blue dye (50mg/kg i.v.) under pentobarbital anaesthesia. Following removal of knee joints and formamide extraction, Evan’s Blue dye extravasation (measured spectrophotomically) was significantly higher in contralateral knee joints of CFA rats than in the knee joints of control rats (p<0.01, 1-way ANOVA). These data indicate that vascular changes had occurred in the uninvolved joint, contralateral to the arthritis. Here we report that a unilateral arthritis results in the generation of contralateral spontaneous antidromic activity in capsaicin-sensitive sensory nerve fibres. In addition, vascular changes (dilatation/increased permeability) were found in knee joints contralateral to arthritis when no overt contralateral inflammation was present. In conclusion, increases in antidromic activity in sensory nerve fibres and vascular changes occur contralateral to CFA-induced arthritis.