During mating, male insects of certain species transfer a costly nuptial gift, a large spermatophore, which is eaten by the female as sperm transfer into her. The spermatophore components (the sperm-free spermatophylax and the sperm ampulla) vary greatly in size between species, and have a direct influence on male fitness. Studies of the relationship between spermatophore size variation and male fitness have concentrated on associations between evolutionary changes in spermatophylax size and either ampulla size or sperm number. Two main hypotheses have been put forward to explain the function of the spermatophylax: the ejaculate-protection hypothesis and the paternal investment hypothesis. A strong correlation between the spermatophylax and ampulla or sperm number suggests an ejaculate-protection function because it protects the ampulla from being removed prematurely. However, comparative support comes mainly from disparate bush-cricket species (Tettigoniidae), that vary greatly in relatedness and diet. Furthermore, data are often from animals reared under laboratory conditions. Our study describes the significance of size variation in bush-cricket nuptial gifts, with an analysis from field populations of 33 species within the genus Poecilimon. Poecilimon share similar diets and the variation in spermatophore size within the genus approximates family-wide variation, so confounding influences from diet and relatedness are, to a certain extent, controlled. Previous support for the ejaculate-protection hypothesis is almost universal, so we expected to find similar results. However, unlike previous studies, there was no relationship between body mass and each of the three spermatophore components when body mass was accounted for, or between spermatophylax mass and sperm number. We also found only a weak relationship between ampulla mass and sperm number, suggesting that caution is needed when using ampulla size to predict sperm number or sperm number to predict ejaculate size. In support of the ejaculate-protection hypothesis we found a positive relationship between spermatophylax size and ampulla mass. While our results support the ejaculate-protection hypothesis, they are not inconsistent with the paternal investment hypothesis.