It has been hypothesized that carotenoid-based sexual ornamentation signals male fertility and sperm competitive ability as both ornamentation and sperm traits may be co-affected by oxidative stress, resulting in positive covariation (the 'redox-based phenotype-linked fertility hypothesis'; redox-based PLFH). On the other hand, the 'sperm competition theory' (SCT) predicts a trade-off between precopulatory and postcopulatory traits. Here, we manipulate oxidative status (using diquat dibromide) and carotenoid availability in adult zebra finch (Taeniopygia guttata) males in order to test whether carotenoid-based beak ornamentation signals, or is traded off against, sperm resistance to oxidative challenge. Initial beak colouration, but not its change during the experiment, was associated with effect of oxidative challenge on sperm velocity, such that more intense colouration predicted an increase in sperm velocity under control conditions but a decline under oxidative challenge. This suggests a long-term trade-off between ornament expression and sperm resistance to oxidative challenge. Shortening of the sperm midpiece following oxidative challenge further suggests that redox homeostasis may constrain sperm morphometry. Carotenoid supplementation resulted in fewer sperm abnormalities but had no effect on other sperm traits. Overall, our data challenge the redox-based PLFH, partially support the SCT and highlight the importance of carotenoids for normal sperm morphology.