Ingestion of antioxidants has been argued to scavenge circulating reactive molecules (e.g., free radicals), play a part in mate choice (by mediating access to this important resource), and perhaps increase life span. However, recent work has come to question these relationships. We have shown elsewhere in the polychromatic lizard, Ctenophorus pictus, that diet supplementation of carotenoids as antioxidants does not depress circulating natural reactive oxygen species (ROS) levels and leads to no corresponding improvement of color traits. However, a much stronger test would be to experimentally manipulate the ROS levels themselves and assess carotenoid-induced ROS depression. Here, we achieve this by using carbonyl cyanide 3-chlorophenylhydrazone, which elevates superoxide (SO) formation approximately threefold at 10 mu M in this model system. We then look for depressing effects on ROS of the carotenoids in order to assess whether 'super-production' of SO makes carotenoid effects on elevated ROS levels detectable. The rationale for this treatment was that if not even such elevated levels of SO are reduced by carotenoid supplementation, the putative link carotenoids, ROS depression, and mate quality (in terms of antioxidant capacity) is highly questionable. We conclude that there is no significant effect of carotenoids on mean SO levels even at the induced ROS levels. However, our results showed a significant interaction effect between carotenoid treatment and male color, with red males having higher ROS levels than yellow males. We suggest that this may be because different pigments are differently involved in the generation of the integumental colors in the two morphs with concomitant effects on ROS depletion depending on carotenoid uptake or allocation to coloration and antioxidation.