BACKGROUND: Anti-tumor necrosis factor (TNF) agents are effective in treating patients with rheumatoid arthritis (RA), but they are associated with (dose-dependent) adverse effects and high costs. To prevent overtreatment, several trials have assessed the effectiveness of down-titration compared with continuation of the standard dose. OBJECTIVES: To evaluate the benefits and harms of down-titration (dose reduction, discontinuation or disease activity guided dose tapering) of anti-TNF agents (adalimumab, certolizumab pegol, etanercept, golimumab, infliximab) on disease activity, functioning, costs, safety and radiographic damage compared with usual care in patients with RA and low disease activity. SEARCH METHODS: We searched the Cochrane Central Register of Controlled Trials (CENTRAL), Issue 8, 2013; Ovid MEDLINE (1946 to 8 September 2013); EMBASE (1947 to 8 September 2013); Science Citation Index (Web of Science); and conference proceedings of the American College of Rheumatology (2005 to 2012) and European League against Rheumatism (2005 to 2013). We contacted authors of the seven included studies to ask for additional information on their study; five responded. SELECTION CRITERIA: Randomised controlled trials (RCTs) and controlled clinical trials (CCTs) comparing down-titration (dose reduction, discontinuation, disease activity-guided dose tapering) of anti-TNF agents (adalimumab, certolizumab pegol, etanercept, golimumab, infliximab) to usual care/no down-titration in patients with RA and a low disease activity state. DATA COLLECTION AND ANALYSIS: Two review authors independently selected studies, assessed risk of bias and extracted data. MAIN RESULTS: Six RCTs and one CCT (total 1203 participants), reporting anti-TNF down-titration, were included. Three studies (559 participants) reported anti-TNF dose reduction compared with anti-TNF continuation. Five studies (732 participants) reported anti-TNF discontinuation compared with anti-TNF continuation (two studies assessed both anti-TNF discontinuation and dose reduction), and one study assessed disease activity-guided anti-TNF dose tapering (137 participants). These studies include only adalimumab and etanercept; controlled data on other anti-TNF agents are absent. Two studies were available in full text; one was assessed as having low risk of bias and the other high risk. Five studies were available only as one or more abstracts. Because data provided in these abstracts were limited, risk of bias was unclear. Clinical heterogeneity between the trials was high.Dose reduction of anti-TNF (etanercept data only) showed no statistically significant or clinical relevant difference in disease activity score in 28 joints (DAS28) (mean difference (MD) 0.10, 95% confidence interval (CI) -0.11 to 0.31) (scale 0.9 to 8; higher score indicates worse disease activity). The proportion of participants who maintained low disease activity was slightly lower among participants given reduced doses of the anti-TNF agent (risk ratio (RR) 0.87, 95% CI 0.78 to 0.98, absolute risk difference (ARD) 9%). Radiographic outcome was slightly worse, but this was not clinically meaningful, compared with continuation of anti-TNF (MD 0.11, 95% CI 0.08 to 0.14) (scale 0 to 448; higher score indicates greater joint damage). Function was not statistically different between anti-TNF dose reduction and continuation (MD 0.10, 95% CI 0.00 to 0.20) (scale 0 to 3; higher score indicates worse functioning). Reinstalment of anti-TNF after failure of dose reduction showed a 5% risk of persistent flare. Data on numbers of serious adverse events (SAEs) (RR 0.58, 95% CI 0.23 to 1.45, ARD -2%) and withdrawals due to adverse events (AEs) (RR 0.57, 95% CI 0.17 to 1,92, ARD -1%) were inconclusive. Most outcomes were based on moderate quality evidence.Participants who discontinued anti-TNF (adalimumab and etanercept data) had higher mean DAS28 (DAS28-erythrocyte sedimentation rate (ESR): MD 1.10, 95% CI 0.86 to 1.34) and DAS28-C-reactive protein (CRP): MD 0.57 95% CI -0.09 to 1.23) and were less likely to maintain a low disease activity state (RR 0.43, 95% CI 0.27 to 0.68, ARD 40%). Also, radiographic and functional outcomes are worse after anti-TNF discontinuation (MD 0.66, 95% CI 0.63 to 0.69, and MD 0.30, 95% CI 0.19 to 0.41, respectively). Data on numbers of SAEs (RR 1.26, 95% CI 0.61 to 2.63, ARD 2%) and withdrawals due to AEs (RR 0.72, 95% CI 0.23 to 2.24, ARD -1%) were inconclusive. Most outcomes were based on moderate quality evidence.The one study comparing disease activity-guided anti-TNF dose tapering (adalimumab and etanercept data) reported no statistically significant differences in functional outcomes (MD 0.20, 95% CI -0.02 to 0.42). Significantly higher mean disease activity was found among participants with tapered anti-TNF at study end (MD 0.50, 95% CI 0.11 to 0.89). No full text of this trial was available for this review. No other major outcomes were reported. All outcomes were based on low quality evidence. AUTHORS' CONCLUSIONS: We can conclude, mostly based on moderate quality evidence, that non-disease activity guided dose reduction of etanercept 50 mg weekly to 25 mg weekly, after at least three to 12 months of low disease activity, seems as effective as continuing the standard dose with respect to disease activity and functional outcomes, although dose reduction significantly induces minimal and not clinically meaningful differences in radiological progression. Discontinuation (also without disease activity-guided adaptation) of adalimumab and etanercept is inferior to continuation of treatment with respect to disease activity and radiological outcomes and function. Disease activity-guided dose tapering of adalimumab and etanercept seems slightly inferior to continuation of treatment with respect to disease activity, with no difference in function. However the only study investigating this comparison included lower than projected numbers of participants.Caveats of this review are that available data are limited. Also, the heterogeneity between studies and the suboptimal design choices (including absence of disease activity-guided dose reduction and discontinuation and use of superiority designs) limit definitive conclusions. None of the included studies assessed long-term safety and costs, although these factors are specific reasons why clinicians consider lowering the dose or stopping the administration of anti-TNF agents.Future research should include other anti-TNF agents; assessment of disease activity, function and radiographic outcomes after longer follow-up; and assessment of long-term safety, cost-effectiveness and predictors for successful down-titration. Also use of a validated flare criterion, non-inferiority designs and disease activity-guided instead of fixed-dose tapering or stopping would allow researchers to better interpret study findings and generalise the information to clinical practice.