Studies were conducted to develop a pheromone‐trap monitoring system for the sweet potato weevil, Cylas formicarius (Fabricius), in sweet potato,Ipomoea batatas(L.) Lam., fields in four states: Florida, Louisiana, North Carolina, and Texas. The present studies examined the attractiveness of sex pheromone synthesized by two sources (USDA and AgriSense) and pheromone of different purities (75–99%) to weevils in the field. In all but one trial, weevil counts did not differ between traps baited with pheromone from the two sources for each of two doses (10 and 100 μg).Percentages of weevils caught per replicate per sample date were not consistently positively correlated with purity (significant in four of 10 trials); however, slope estimates were steeper in regions with lower trap counts (North Carolina and Louisiana) than in those with higher trap counts (Florida). Because the pheromone will be used to monitor weevils In both weevil‐free and weevil‐infested regions, the use of pheromone with a purity level >99% is most appropriate. The importance of these data in pheromone‐trap monitoring programmes for this weevil worldwide is discussed.