Objective: Studies have showed that different follow-up starting points might potentially impact the comparison between primary (PMIBC) and secondary muscle-invasive bladder cancer (SMIBC), but the only previous meta-analysis did not differentiate the follow-up starting points of included studies. With more trials published, we aim to update the meta-analysis comparing PMIBC and SMIBC. Methods: PubMed, Embase, Cochrane Library and ClinicalTrial.gov. systematically searched. Literatures comparing the survival outcomes of PMIBC and SMIBC were selected. Outcomes of cancer-specific mortality (CSM), overall mortality (OM) and recurrence-free survival (RFS) were pooled and grouped based on the starting point of follow-up (after initial diagnosis or radical cystectomy (RC)). Newcastle-Ottawa Scale (NOS) and funnel plot were employed to assess the study quality and publication bias, respectively. Results: A total of 17 high-quality studies were selected, with 5558 patients aged from 59.8 to 72.7 (mean value) involved. The male-to-female ratio was roughly 4:1 (4390/1124). SMIBC had lower risk of CSM after initial diagnosis (HR 0.81, 95%CI 0.67-0.98, P=0.03, I2=70%), but higher risk of CSM after RC (HR 1.45, 95%CI 1.27-1.65, P<0.00001, I2=64%). In terms of OM and recurrence, outcomes were pooled only after RC, which both turned out to be higher for SMIBC (OM: HR 1.50, 95%CI 1.30-1.73, P<0.00001, I2=0%; Recurrence: HR 1.66, 95%CI 1.36-2.02, P<0.00001, I2=48%). No obvious publication bias was observed from funnel plot. Conclusion: The current study suggested SMIBC had higher risk of CSM, OM and recurrence after RC, but lower risk of CSM after initial diagnosis. © The author(s).