Three suitable sex pheromone trap designs for trapping the red bofiworm of cotton, Diparopsis castanea Hmps., including a simple omnidirectional monitoring trap, are described. In trials in Malawi, Diparopsis was more readily trapped on open horizontal trapping surfaces (water or adhesive) than in enclosed types of traps. The presence of a roof above the trapping surface, at spacings of up to 20 cm, significantly reduced catches. White traps, or unpainted traps in the case of vane traps constructed from galvanised metal, caught more moths than traps of other colours. Catches were greatest when monitoring traps in cotton were placed at ground level or at a height of 0·5 m and, when dicastalure was released from polyethylene vials, by employing a trap 60 cm in diameter. Evidence was obtained that competition may exist between female-baited traps spaced up to 73 m apart, although catches in traps baited with dicastalure were not significantly different when traps were 40–60 m apart. The greatest catches of moths in cotton were in traps located in insecticide-sprayed fields, but moths were also trapped in considerable numbers in other crops and to a reduced extent in open areas 4·8 km from the nearest cotton. From about the end of February, catches in maize, groundnut and fallow areas frequently exceeded the catches recorded within cotton fields, which may indicate an increased dispersal flight from cotton by adult moths later in the season. Weekly replacement in monitoring traps of polyethylene vials containing dicastalure produced maximum moth catches, but there was no significant difference in catch between traps in which pheromone was replaced at intervals of 2, 3, 4 or 6 weeks.