Background: EVE and AXI are approved as 2^nd TTs for mRCC. This study compares OS and PFS among mRCC patients (pts) treated with EVE and AXI following 1^st TKI. The extent to which duration of 1^st TKI treatment modifies comparative effectiveness of 2^nd TT is also assessed. Methods: Retrospective reviews of medical records were conducted by medical oncologists or hematologists/oncologists recruited from a nationwide panel. Pt eligibility criteria included: 1) aged ≥ 18 years; 2) initiated and discontinued 1^st TKI (sunitinib [SUN], sorafenib [SOR], or pazopanib [PAZ]) for medical reasons; 3) initiated 2^nd TT between 2/2012 and 1/2013. OS was defined as time from initiation of 2^nd TT to death. PFS was defined as time from initiation of 2^nd TT to physician-assessed progression or death, whichever occurred first. Multivariable Cox proportional hazards models were used to estimate the comparative hazard ratios (HRs) for OS and PFS between EVE and AXI, adjusting for age, gender, type and duration of 1^st TKI, response to 1^st TKI, duration of mRCC at 2^nd TT, metastatic disease at initial diagnosis, clear cell RCC, prior nephrectomy, performance status, metastatic sites, comorbidity, and years of physician practice. Comparative effectiveness was also analyzed by type and duration ( < 6, 6-12, ≥ 12 months) of 1^st TKI. Results: 325 and 127 pts received 2^nd TT with EVE and AXI. After adjusting for baseline characteristics, there was no statistically significant difference between EVE and AXI for OS [HR (95% CI): 1.16 (0.73-1.82)] or PFS [HR (95% CI): 1.16 (0.85-1.59)]. When stratified by type and duration of 1^st TKI, there was no statistically significant difference in OS between EVE and AXI in all subgroups, except for pts with < 6 months on SUN or SOR as 1^st TT [HR (95% CI): 2.98 (1.10, 8.12)]. No statistically significant difference in PFS was observed in any subgroup. Conclusions: In this large, retrospective chart review, there was no significant difference in OS or PFS between EVE and AXI in the overall population. Longer durations of 1^st TKI were not associated with better comparative effectiveness for subsequent treatment with AXI vs. EVE.