Daily catches of gamma sterilised male Glossina austeni Newstead with experimental sticky panels were analysed from March 1996 to July 1997. The flies were released weekly by light aircraft over primary and secondary forest ecosystems of Unguja Island, Zanzibar. In the primary forest, the cross-shaped royal blue XT panel (two interlocking panels of each 70 x 60 cm) trapped significantly more flies than the royal blue-white leg panel (panel with a body of 65 x 30 cm and two legs of each 15 x 15 cm) in all months, except in July 1996 and July 1997. In the same habitat, the cross-shaped royal blue-white leg panel (two interlocking leg panels) trapped from 1.7 (not significant) to 3.0 times (highly significant) as many flies as the standard leg panel depending on the season. Significantly more flies were trapped with the cross-shaped XT panel than with the leg panel in the secondary forest compared to the primary forest. Catches of the cross-shaped XT panel on each of a series of days, correlated well with those of the leg panel on the same day, except for panels deployed in the primary forest during the hot-dry and the beginning of the cold-dry season. Catches of the cross-shaped leg panel correlated less well with the catches of the standard leg panel in the primary forest. The data presented indicate that the behavioural responses in time and space of sterile male G. austeni are influenced by the type of trapping device used.