Bat surveys are frequently undertaken using ultrasonic detectors to determine the species present in an area on the basis of the identity of echolocation calls. We compared three techniques for using the Anabat II detector: the detector pointed along tracks (flyways) versus the detector pointed across tracks (non-flyways); recording output to audio cassette (analogue) versus direct recording to computer (digital); and active hand-held recording versus static automatic recording. In addition, we derived a species-accumulation curve from all-night Anabat recordings in the wet–dry tropics of the Northern Territory. We found no significant difference between flyway and non-flyway recordings; significantly more calls were identified from digital recordings; and significantly more species were detected using hand-held than static recordings. Species-accumulation analysis suggests that the minimum time required to achieve a satisfactory (80%) inventory of bat species at a site is during the three-hour period immediately after sunset. We use our findings to make recommendations for the design of bat surveys using the Anabat II detector.