A grid of 25 traps baited with virgin female moths was used to monitor the abundance of male spruce budworm in relatively sparse populations for 12 successive years. Concurrent sampling of third-instar larvae showed that the male moth counts of one generation had a predictive relationship to third-instar larval counts of the next generation and therefore the sex-attractant trap is a feasible means of measuring budworm abundance at low population densities. But three sampling constraints must be noted: (a) a moth invasion will generally invalidate the relationship;(b)extrapolation of a moth–larvae relationship observed in one forest biotype to biotypes with different host characteristics is risky; (c) the probability of a female attracting a particular male decreases sharply with increasing population density.Trapping success is discussed with respect to location and height within the stand and to weather conditions.