Fire-favoured insects are difficult to sample except opportunistically after forest fires. Here, we tested if smoke from a small fire could be an efficient way to sample such insects. Insects were sampled over ca. 10 h hours, by hand-picking and netting on screens put up around the fire. Two specimens of the rare and redlisted Hormopeza spp. (Diptera, Empididae) were caught. Large numbers ([20,000) of Microsania spp. (Diptera, Platypezidae) were caught, but none in the absence of smoke. The numbers of Microsania spp. clearly peaked in late afternoon, and a short sampling period would be sufficient if targeting only this taxon. Of the almost 200 species of Coleoptera, 17 % were considered as fire-favoured, contributing 9 % of the specimens, suggesting low efficiency of the method for this group. Using 23 sites differing in fire history, catches of Microsania spp. were unaffected by numbers and area of forest fire (preceding 5 years and within 10 km radius over the sampling sites). In contrast, there was a weak trend for the proportion of fire-favoured Coleoptera to increase with increasing number of fires. To conclude, smoke as produced in our study can clearly attract fire-favoured Diptera, but smoke had only a weak effect on fire-favoured Coleoptera in the study area. It is still likely that selectively picking specimens of species attracted to smoke is a more cost-efficient method than using, e.g., Malaise traps that catch indiscriminately.