A previous meta-analysis showed that quinolones administered for prophylaxis of infections among cancer patients reduced all-cause mortality. We extracted from the primary trials infection-related and all-cause mortality as reported and assessed the effect of quinolones on non-infection-related mortality through meta-analysis. Among trials comparing quinolones to placebo or no treatment, a significant reduction in non-infection-related mortality was observed (relative risk 0.54, 95% confidence interval 0.32-0.93, 15 trials, 3,320 patients). This finding might represent biased attribution of deaths to infection or might be compatible with an anti-cancer effect of quinolone antibiotics. We present further analyses addressing these possibilities.