BackgroundNanoparticle albumin‐bound (nab)‐paclitaxel shows promising activity and tolerability in platinum‐refractory metastatic urothelial cancer (mUC).This phase III trial aimed to compare the efficacy and safety of nab‐paclitaxel plus gemcitabine (GA) with that of carboplatin plus gemcitabine (GCb) as a first‐line treatment for patients with cisplatin‐ineligible mUC.Patients and methodTreatment‐naive, cisplatin‐ineligible patients with mUC were assigned randomly to either the GA (both nab‐paclitaxel 125 mg/m2 and gemcitabine 1,000 mg/m2 on days 1 and 8, every 21 days) or GCb group (carboplatin AUC 4.5 on day 1, gemcitabine 1000 mg/m2 on days 1 and 8, every 21 days). The primary endpoint was progression‐free survival (PFS). Secondary endpoints included objective response rate (ORR), disease control rate (DCR), overall survival (OS), safety and patient‐reported outcomes (PROs).ResultsThe trial was terminated early because of slow accrual after 54 patients were enrolled: 26 in in the GA group and 28 in the GCb groups. The median PFS was 6.7 months versus 5.9 months for the GA and GCb groups, respectively (P=0.248). The median OS time was 12.1 versus 10.7 months for the GA and GCb groups, respectively (P=0.837). The ORR and DCR were 40% versus 46.4% (P=0.637) and 72% versus 68% (P=0.188) in the GA and GCb groups, respectively.Patients treated with GA showed significantly lower incidence of grade 3‐4 thrombocytopenia, and does reduction and delay. Although peripheral sensory neuropathy was higher in GA arm, no grade 3 neuropathy occurred. There was no difference in the PROs bwteen the two groups.ConclusionWhile not powered for comparison, first‐line GA showed similar efficacy and better tolerability with GA, and might be considered a rational alternative to GCb.