Objective To investigate the associations of a range of personal and clinical variables with bladder cancer survival in men and women in NSW to see if we could explain why bladder cancer survival is consistently poorer in women than in men. Patients and Methods All 6880 cases of bladder cancers diagnosed in NSW between 2000 and 2008 were linked to hospital separation data and to deaths. Separate Cox proportional hazards regression models of hazard of bladder cancer death were constructed for those who did or did not undergo cystectomy. Results A total of 16% of patients with bladder cancer underwent cystectomy (16% of men and 15% of women). Women who underwent cystectomy were 26% more likely to die than men (hazard ratio [HR] 1.26, 95% confidence interval [CI] 1.00-1.59) after adjustment for age, stage, time from diagnosis to cystectomy, distance from treatment facility and country of birth. None of the above covariates had a material effect on the difference in hazard between women and men; however, when stratified by a history of cystitis, the adjusted hazard was 55% higher in women (HR 1.55, 95% CI 1.15-2.10) than in men with a history of cystitis while, in the absence of this history, there was no difference in the hazard between men and women (HR 0.99, 95% CI 0.57-1.70). This apparent modification of the effect of sex on bladder cancer outcome was not seen in patients treated only by resection: the adjusted HRs in women relative to men were 1.10 (95% CI 0.92-1.31) in those with a history of cystitis and 1.21 (95% CI 0.98-1.50) in those without. A history of haematuria did not modify appreciably the association of sex with bladder cancer outcome. Conclusion Women's poorer survival from bladder cancer compared with that of men remains unexplained; however, the possibility that some factor associated with a history of cystitis may contribute to or explain the poorer outcome in women merits further investigation.