Background: Cabozantinib (CABO) is an oral tyrosine kinase inhibitor registered for the treatment of metastatic renal cell carcinoma (mRCC) for the first or subsequent lines. Tolerability in real world elderly patients is poorly documented. G8 is a short test for vulnerability gaining increased interest as a screening tool for trials in geriatric oncology. Methods: ZEBRA/MEET-URO 9 was a prospective multicenter study of safety and activity of CABO administered to pts ≥70 years with mRCC, either in the first or subsequent lines of treatment, until progression or unacceptable toxicity. All pts underwent G8 score at baseline, with a cut-off for vulnerability of 14 or below. Data on tolerability and activity were collected prospectively after signature of informed consent. Results: A total of 104 pts started CABO at 13 Italian Centers, 38,5% as first line. Median age was 75.8 yrs (range 70.2-87.4 yrs, 26 pts ≥80 yrs), 73.1% males. IMDC score was good 19.2%, intermediate 53.9%, poor 26.9%. Primary tumor had been removed in 82.7% of pts, histology was clear cell 78.8%, papillary 8.7%, chromophobe 5.8%, unclassified 6.7%. G8 score was ≤14 in 65.4% of pts. Up-front dose reduction of CABO was more frequent in pts with low G8 score (79.4 vs 41.7%, p=0.003), but eventually the majority of pts (91.4%) underwent dose reductions of CABO. After a median treatment of 6.4 months (0.5-26.1 months), 38.4% of pts developed G3-4 toxicities, 22.1% interrupted treatment due to adverse events, 2.8% (3 pts) died due to cardiovascular or thromboembolic events. Median PFS was 7.6 months (95% CI=5.8-12.6 months) in first line, 10.0 months (5.8-15.6) in second or further lines, median OS was 20.1 months (11.1-not reached) and 15.6 months (12.5-not reached), respectively. G8 score ≤14 did not correlate with rate of temporary interruptions >7 days, hospitalization, incidence of G3-5 toxicities, as well as with PFS. Pts with G8 score ≤14 had a trend for reduced OS, but difference was not statistically significant both in the first and further lines of treatment. Conclusions: Screening G8 test was positive in more than a half of pts, underlying the need for detailed geriatric assessment and increased clinical monitoring of such patients. A G8 score ≤14 correlated with up-front dose reduction of CABO but not with G3-5 toxicities probably due to the high rates of dose reductions in the whole cohort. Correlation between low G8 score and OS could not be demonstrated in this population.