Morphological structures used as weapons in male–male competition are not only costly to develop but are also probably costly to maintain during adulthood. Therefore, having weapons could reduce the energy available for other fitness‐enhancing actions, such as post‐copulatory investment. We tested the hypothesis that armed males make lower post‐copulatory investments than unarmed males, and that this difference will be most pronounced under food‐limited conditions. We performed two experiments using the male‐dimorphic bulb mite Rhizoglyphus robini, in which males are either armed “fighters” or unarmed “scramblers.” Firstly, we tested whether fighters and scramblers differed in their reproductive output after being starved or fed for 1 or 2 weeks. Secondly, we measured the reproductive output of scramblers and fighters (starved or fed) after one, two or three consecutive matings. Scramblers sired more offspring than fighters after 1 week, but scramblers and fighters only sired a few offspring after 2 weeks. Scramblers also sired more offspring than fighters at the first mating, and males rarely sired offspring after consecutive matings. Contrary to our hypothesis, the fecundity of starved and fed males did not differ. The higher reproductive output of scramblers suggests that, regardless of nutritional state, scramblers make larger post‐copulatory investments than fighters. Alternatively, (cryptic) female choice generally favours scramblers. Why the morphs differed in their reproductive output is unclear. Neither morph performed well relatively late in life or after multiple matings. It remains to be investigated to what extent the apparent scrambler advantage contributes to the maintenance and evolution of male morph expression.