Indirect genetic effects (IGEs) occur when genes expressed in one individual alter the phenotype of an interacting partner. IGEs can dramatically affect the expression and evolution of social traits. However, the interacting phenotype(s) through which they are transmitted are often unknown, or cryptic, and their detection would enhance our ability to accurately predict evolutionary change. To illustrate this challenge and possible solutions to it, we assayed male leg‐tapping behavior using inbred lines of Drosophila melanogaster paired with a common focal male strain. The expression of tapping in focal males was dependent on the genotype of their interacting partner, but this strong IGE was cryptic. Using a multiple‐regression approach, we identified male startle response as a candidate interacting phenotype: the longer it took interacting males to settle after being startled, the less focal males tapped them. A genome‐wide association analysis identified approximately a dozen candidate protein‐coding genes potentially underlying the IGE, of which the most significant was slowpoke. Our methodological framework provides information about candidate phenotypes and candidate single‐nucleotide polymorphisms that underpin a strong yet cryptic IGE. We discuss how this approach can facilitate the detection of cryptic IGEs contributing to unusual evolutionary dynamics in other study systems.