The prognostic relevance of primary location of urothelial carcinoma on survival has been poorly investigated. We used prospectively collected data from 3 European Organization for the Research and Treatment of Cancer advanced urothelial carcinoma studies, including 30924 (methotrexate, vinblastine, doxorubicin and cisplatin vs high dose methotrexate, vinblastine, doxorubicin and cisplatin), 30986 (methotrexate, carboplatin and vinblastine vs gemcitabine and cisplatin in patients who were not candidates for cisplatin) and 30987 (gemcitabine and cisplatin-paclitaxel vs gemcitabine and cisplatin in candidates for cisplatin). Patients were grouped by primary tumor location as those with bladder cancer or upper tract urothelial carcinoma. Progression-free and overall survival was tested by tumor location using Cox proportional hazard regression stratified by study and treatment using 2-sided α = 0.05. Of the 1,039 patients 878 (85.3%) and 161 (14.7%) had bladder cancer and upper tract urothelial carcinoma, respectively. Patients with bladder cancer had better performance status and were more likely to be male (p = 0.008 and <0.074, respectively). By a median followup of 4.8 years (IQR 4.0-6.7) 733 patients had died and 925 had experienced disease progression. Overall and progression-free survival did not differ significantly between bladder and upper tract urothelial carcinoma cases (p = 0.3 and 0.7, respectively), even after adjusting for the effects of Bajorin risk group by each tumor location. When upper tract urothelial carcinoma was considered separately, patients with primary ureteral tumors had better overall survival than patients with primary bladder cancer (OR = 0.74, p = 0.047). However, this association did not remain significant after adjusting for Bajorin risk group (p = 0.05). Primary tumor location had no impact on progression-free or overall survival in patients with locally advanced or metastatic urothelial carcinoma treated with platinum based combination chemotherapy.