1. There is increasing evidence to show that the environment experienced by fathers can have intergenerational paternal effects on the phenotype of offspring. Paternal effects can be mediated by ejaculate traits beyond the genes carried by sperm, for example through methylation of DNA during spermatogenesis, or non-sperm components of the ejaculate such as small RNAs or seminal fluid proteins (sfp).

2. Theory suggests that ejaculate-mediated paternal effects can have important evolutionary consequences, especially if they exhibit condition dependent expression. Here we use the nutritional geometry approach to explore condition dependence in ejaculate traits and intergenerational paternal effects in the cricket Teleogryllus oceanicus.

3. We fed males one of 24 chemically defined diets that varied in macronutrient concentration and the ratio of protein (P) to carbohydrate (C). We measured paternal expenditure on testes and accessory gland growth, and the expression of three seminal fluid protein genes that are known to impact sperm viability and the survival of embryos to hatching. We also measured the hatching success and reproductive phenotypes of adult offspring.

4. We found linear effects of P and C on testes and accessory gland mass, and linear and correlated effects of P and C on the expression of two sfp genes. Paternal diet had no effect on the survival of embryos to hatching.

5. There were significant intergenerational effects of paternal diet on the viability of sperm produced by sons, but no paternal effects on the expression of sfp genes in sons or fecundity in daughters.

6. Our data suggest that ejaculate quality can have a significant non-genetic component determined by the paternal environment, which holds important implications for the evolutionary maintenance of variation in heritable male fitness.