This study investigated the effects of different artificial insemination (AI) regimes on the pregnancy rate in mares inseminated with either cooled or frozen-thawed semen. In essence, the influence of three different factors on fertility was examined; namely the number of inseminations per oestrus, the time interval between inseminations within an oestrus, and the proximity of insemination to ovulation. In the first experiment, 401 warmblood mares were inseminated one to three times in an oestrus with either cooled (500×10 6 progressively motile spermatozoa, stored at +5 °C for 2–4 h) or frozen-thawed (800×10 6 spermatozoa, of which ≥35% were progressively motile post-thaw) semen from fertile Hanoverian stallions, beginning −24, −12, 0, 12, 24 or 36 h after human chorionic gonadotrophin (hCG) administration. Mares were injected intravenously with 1500 IU hCG when they were in oestrus and had a pre-ovulatory follicle ≥40 mm in diameter. Experiment 2 was a retrospective analysis of the breeding records of 2637 mares inseminated in a total of 5305 oestrous cycles during the 1999 breeding season. In Experiment 1, follicle development was monitored by transrectal ultrasonographic examination of the ovaries every 12 h until ovulation, and pregnancy detection was performed sonographically 16–18 days after ovulation. In Experiment 2, insemination data were analysed with respect to the number of live foals registered the following year. In Experiment 1, ovulation occurred within 48 h of hCG administration in 97.5% (391/401) of mares and the interval between hCG treatment and ovulation was significantly shorter in the second half of the breeding season (May–July) than in the first (March–April, P≤0.05). Mares inseminated with cooled stallion semen once during an oestrus had pregnancy rates comparable to those attained in mares inseminated on two (48/85, 56.5%) or three (20/28, 71.4%) occasions at 24 h intervals, as long as insemination was performed between 24 h before and 12 h after ovulation (78/140, 55.7%). Similarly, a single frozen-thawed semen insemination between 12 h before (31/75, 41.3%) and 12 h after (24/48, 50%) ovulation produced similar pregnancy rates to those attained when mares were inseminated either two (31/62, 50%) or three (3/9, 33.3%) times at 24 h intervals. In the retrospective study (Experiment 2), mares inseminated with cooled semen only once per cycle had significantly lower per cycle foaling rates (507/1622, 31.2%) than mares inseminated two (791/1905, 41.5%), three (464/1064, 43.6%) or ≥4 times (314/714, 43.9%) in an oestrus ( P≤0.001). In addition, there was a tendency for per cycle foaling rates to increase when mares were inseminated daily (619/1374, 45.5%) rather than every other day (836/2004, 42.1%, P=0.054) until ovulation. It is concluded that under conditions of frequent veterinary examination, a single insemination per cycle produces pregnancy rates as good as multiple insemination, as long as it is performed between 24 h before and 12 h after AI for cooled semen, or 12 h before and 12 h after AI for frozen-thawed semen. If frequent scanning is not possible, fertility appears to be optimised by repeating AI on a daily basis.