Aim: To determine whether exposure to tumour necrosis factor (TNF)-a inhibitors increases the risk of herpes zoster (HZ) among people with rheumatoid arthritis (RA). Methods: We performed a cohort study of people with RA participating in the Australian Rheumatology Association Database. We identified self-reported cases of HZ and verified using medical records. For the primary analysis, we only included doctor-verified cases. For TNF-a inhibitor exposed groups, we excluded HZ episodes that occurred before TNF-a inhibitor initiation, and for the control group we excluded HZ episodes that occurred prior to 2000 or RA diagnosis. The risk of HZ among participants exposed versus not exposed to TNF-a inhibitors was compared using Cox proportional hazards models including significant covariates affecting the risk. Adjusted hazard ratios (HR) were calculated for TNF inhibitors as a class and for individual agents. Results: Among 2157 active RA participants, there were 442 self-reported cases of HZ. From 346 responses from doctors, 249 cases were verified and four were false positives (false positive rate 1.6 ). Crude incidence of verified HZ in the entire RA cohort was 15.9/1000 person-years (95 confidence interval (CI): 13.5-18.8). An increased risk of HZ was found for all TNF-a inhibitors combined (fully adjusted HR 1.71; 95 CI: 1.00-2.92) and adalimumab (fully adjusted HR 2.33; 95 CI: 1.22-4.45), but in the fully adjusted model was not increased with etanercept (fully adjusted HR 1.65; 95 CI: 0.90-3.03). No increased risk was found with infliximab (HR 1.29; 95 CI: 0.37-4.47). Conclusions: TNF-a inhibitors are associated with an increased risk of HZ in people with RA compared with those who have not been exposed.