Neither a small nor a large population of Anthonomus grandis Boheman was suppressed by from 10–80 pheromone traps/ha though the optimum number of traps/ha (greatest number of weevils captured/trap) was determined to be 20 for a large population. The regression of square infestation on trap capture was best expressed by the equation y = 0.09+0.048X. During the 1st and 2nd week of trapping (prior to squaring), captures within the field were clumped and fit a negative binominal distribution. Eighty % of the total weevils were captured and 75% of the punctured squares were found in a 20-m-wide belt around a cotton field. Boll weevil pheromone traps therefore can function alone only as a survey and detection tool; traps must be combined with other techniques when used in an integrated control program.