Weapons are among the most extravagant sexually selected traits known, yet the evolution of weapon diversity remains understudied. We used field crickets (Orthoptera, Gryllinae) to test two hypotheses explaining interspecific diversity in weaponry. We raised eight species ofGryllusfield crickets under common garden conditions and staged interactions between conspecific males. We measured body size and weapon shape (relative head and mouthpart size) to determine weapon allometry in both males and females, quantified the intensity of male-male aggression for each species, analyzed the effects of both body size and weapon shape on contest outcome, and tested comparative relationships between morphology and behaviour using phylogenetic least squares regression. We found that larger males won more contests than smaller males in seven of eight species, and weapon shape predicted contest success in only one species. Contrary to the fighting advantage hypothesis, body size was not related to aggressiveness across species, but weaponry was. Additionally, the most aggressive species had the most elaborate weaponry, contrary to the weapon-signal continuum hypothesis. Our results highlight the complexity of weaponry evolution in a group of organisms that has been a model system for the observation and study of aggression for approximately 1000 years.