OBJECTIVES: Several therapies have recently been approved for routine commissioning in the UK National Health Service (NHS) for pre-treated amRCC, but comparative evidence to guide decisions between them is limited. This research evaluated the clinical effectiveness of axitinib, cabozantinib, everolimus, nivolumab, sunitinib and BSC for people with amRCC previously treated with vascular endothelial growth factor (VEGF)-targeted therapy. METHODS: Systematic review and MTC of randomised controlled trials (RCTs) and non-RCTs. Primary outcomes were overall survival (OS) and progression-free survival (PFS). Secondary outcomes were overall response rates (ORR), adverse effects (AEs), and health-related quality of life (HRQoL). MEDLINE, EMBASE, and The Cochrane Library were searched in January and June 2016 for RCTs and non-RCTs, respectively. Two or more reviewers abstracted study data and performed critical appraisals. Fixed-effect MTC was conducted for OS, PFS and ORR. Effect estimates included hazard ratio (HR) and odds ratio with 95% credible intervals (CrI). RCT data formed the primary analyses, with non-RCTs and studies at high risk of bias included in sensitivity analyses. HRQoL and AEs were summarised narratively. RESULTS: Four RCTs (n = 2,618) and eight non-RCTs (n = 1,526) were included. No suitable studies for sunitinib were identified. The inclusion of axitinib was limited to the sensitivity analyses. Results for PFS show a statistically significant benefit for cabozantinib vs everolimus (HR 0.51, 95% CrI: 0.41 to 0.63) and both treatments are more effective than BSC. OS results show a benefit for cabozantinib (HR 0.66, 95% CrI: 0.53 to 0.82) and nivolumab (HR 0.73, 95% CrI: 0.60 to 0.89) vs everolimus. Sensitivity analyses were consistent with the primary analyses. CONCLUSIONS: Evidence from RCTs suggest cabozantinib is likely to be the most effective treatment for PFS and OS, closely followed by nivolumab. All treatments appear to prolong PFS and OS compared with BSC although interpretation of some results is limited by heterogeneity.