Long-term treatment with glucocorticoids (GCs) plays an important role in the management of arthritis patients, although the efficacy/safety balance is unfavorable. Alternatives with less (severe) adverse effects but with good efficacy are needed. Selective GC receptor modulators (SGRMs) are designed to engage the GC receptor with dissociative characteristics: transactivation of genes, which is mainly responsible for unwanted effects, is less strong while trans-repression of genes, reducing inflammation, is maintained. It is expected that SGRMs thus have a better efficacy/safety balance than GCs. A systematic review providing an overview of the evidence in arthritis is lacking. To systematically review the current literature on efficacy and safety of oral SGRMs in comparison to GCs in arthritis. A search was performed in Medline, Embase and the Cochrane Library, from inception dates of databases until May 2017. Experimental studies involving animal arthritis models or human material of arthritis patients, as well as clinical studies in arthritis patients were included, provided they reported original data. All types of arthritis were included. Data was extracted on the SGRM studied and on the GC used as reference standard; the design or setting of the study was extracted as well as the efficacy and safety results. A total of 207 articles was retrieved of which 17 articles were eligible for our analysis. Two studies concerned randomized controlled trials (RCT), five studies were pre-clinical studies using human material, and 10 studies involved pre-clinical animal models (acute and/or chronic arthritis induced in mice or rats). PF-04171327, the only compound investigated in a clinical trial setting, had a better efficacy/safety balance compared to GCs: better clinical anti-inflammatory efficacy and similar safety. Studies assessing both efficacy and safety of SGRMs are scarce. There is limited evidence for dissociation of anti-inflammatory and metabolic effects of the SGRMs studied. Development of many SGRMs is haltered in a preclinical phase. One SGRM showed a better clinical efficacy/safety balance.