In the 56 years since the first description of the chemical structure of a pheromone, there has been extensive research effort by chemical ecologists to identify the composition of pheromones across species. This is particularly so for Lepidoptera, where the female sex pheromones of several hundred species have now been described. Identification of these blends of chemical components allows us to test many hypotheses regarding the evolution of the chemical signals used in mate attraction and mate choice. However, this requires accurate identification all of the components, and yet for many species we are likely missing minor, but critical, components of the blend – meaning we have incomplete descriptions the signal. I surveyed the literature for sex pheromone descriptions of over 500 moth species in an attempt to identify phylogenetic patterns in pheromone complexity. After controlling for research effort and time since most recent analysis (both of which are linked, weakly, to number of described components), I found evidence of phylogenetic signal in pheromone composition in likely well-described species. I then used this information to identify likely species from clades where the pheromone complexity has likely been underestimated, when compared to the known complexity of their close relatives.