To evaluate the oncological feasibility of pure laparoscopic radical nephroureterectomy (p-LRNU) for upper tract urothelial carcinoma (UTUC) compared with conventional LRNU (c-LRNU) using a propensity-adjusted multi-institutional collaboration dataset. Among the 503 UTUC patients who underwent RNU, we identified 219 who underwent c-LRNU (laparoscopic nephrectomy with open bladder cuff resection) and 72 who underwent p-LRNU (dissecting the kidney, ureter, and bladder cuff under complete laparoscopy). We adopted a propensity score (PS) matching method to achieve homogeneity with respect to patient backgrounds. PS matching-adjusted Cox-regression analysis was performed to evaluate the risk factors that influenced oncological outcomes. Sixty-eight p-LRNU and 68 c-LRNU patients were matched. Overall, 51 (37.0%) developed intravesical recurrence (IVR), 21 (15.4%) had disease recurrence, and 20 (14.7%) died. Patients who underwent p-LRNU had a significantly shorter operation time and less blood loss than those who underwent c-LRNU. Although no significant differences in 3-year recurrence-free survival were found between the two methods, atypical recurrence sites were observed in the p-LRNU group, including the brain, sigmoid colon, vagina, and peritoneum. Regarding IVR, the 3-year IVR-free survival rate was 41.8% in the p-LRNU group, which was significantly lower than that in the c-LRNU group (66.6%, p = 0.004). Multivariate analysis demonstrated that a history of bladder cancer, ureteral cancer, and p-LRNU were independent risk factors for subsequent IVR. Although p-LRNU is less invasive, the current technique may increase the incidence of atypical disease recurrence and subsequent IVR due to extravesical and intravesical tumor dissemination.