Sex is a factor influencing development in many insect species, but is not widely studied in forensically important blow flies. If sex influences blow fly development, knowing the sex of a larva from a corpse can increase precision in estimates of that larva's age. The improved prediction of larval age will make estimates of time since death using entomological evidence better. Larvae lack sexually dimorphic morphological characteristics, so sex is not immediately known visually. To generate sexually dimorphic reference growth curves, a subsample must be large enough to ensure enough males and females are present for comparison. Using two entire age Chrysomya megacephala (Fabricius) cohorts, we evaluated the minimum sample number needed to have enough individuals of both sexes for comparison using 95% prediction intervals. Through a simulation of three trials of 1000 random replicates, we determined that a sample size of 19 would prevent any instance of a comparison not occurring because of insufficient sampling from one sex. As the current method for molecular sex determination can be expensive, we also compared how the results of various subsampling percentages compare those of the entire age cohorts. We found that subsampling at least 50% of an entire cohort leads to almost identical results in comparison to the entire age cohort. Together, these findings will help establish guidelines for generating sex-specific reference growth curves. A uniform approach to generating these sex-specific growth curves will lead to more consistency in age estimates made from them.