Oestrus was synchronised in 57 Bos indicus heifers using norgestomet-oestradiol and pregnant mare serum gonadotrophin. Oestrus was detected by observations made at six-hourly intervals, using oestrogen-treated and chin-ball harnessed steers, heatmount detectors, tail-paint and visual observation. Heifers were inseminated once at either a fixed time of 49.2 +/- 0.4 h (mean +/- SE; n = 29) after implant removal or 12.6 +/- 1.5 h (n = 28) after oestrus was detected. The mean (+/- SE) time to the onset of oestrus was 47.1 +/- 1.9 h, while 90% of heifers recorded in oestrus were detected within 66 h of implant removal. Heatmount detectors were significantly more efficient at detecting oestrus than chin-ball harnessed steers, tail paint or visual observation (P < 0.001). A higher pregnancy rate was obtained in heifers inseminated after oestrus detection compared with heifers inseminated at a fixed-time (57.1 vs 34.5%; P = 0.043) and a higher pregnancy rate was obtained in heifers classified as easy to inseminate compared with heifers classified as difficult to inseminate (57.8 vs 0%, P < 0.001). We conclude that heatmount detectors are an efficient means of detecting oestrus in synchronised B indicus heifers and that pregnancy rates can be increased when insemination follows oestrus detection compared with a fixed-time insemination regimen.