Evaluation of species diversity for highly diverse taxa is extremely time-consuming and costly; thus, there is a need to develop efficient sampling strategies. We established a short-term, efficient sampling scheme to produce samples that represent a full-season sampling of moth assemblages with a high degree of seasonality. We sampled adult moths monthly for the duration of the moth flying season by using light traps in five forest stands in a cool-temperate region in central Hokkaido, northern Japan. From this sample, we generated various subsamples that reduced the sampling period and the number of traps per stand, and tested whether these subsamples provide estimates of species richness, abundance, and species turnover representative of those revealed by the whole season sample. Comparisons between the whole season sample and each subsample indicated that setting one light trap on a night in July and August, which shortened the sampling period to 25% and reduced sample size to 38%, was the most efficient sampling scheme to estimate abundance, species richness, and similarity in the whole season sample. The comparisons also suggest that it is efficient to use rarified species richness as a species richness estimator, and the Bray—Curtis index or Morisita's Cλ for estimating species turnover between sites in moth assemblages.