Pig producers have been shown keen interest of the number of spermatozoa in a semen dose since pig artificial insemination introduce. However, determining the minimal number of spermatozoa need per AI without detrimental effect on overall reproductive performances is not an easy question to answer. To increase the efficiency of semen utilization in pig AI, optimum number of spermatozoa per dose needed to determine. The objective of this study was to determine the reproductive performance and factors that affect on-farm application of low-dose semen insemination in sows. Data were collected from Darby Genetics AI studs from 4th of June to 7th of July, 2012 (n=401). The numbers of parturition were 84, 234 and 83 in sows inseminated with doses of 1.5×109, 2.0×109 and 2.5×109 spermatozoa in 100ml extender, respectively. There were no significant differences on reproductive performances such as gestation pe-riod, total born, total born alive, stillbirth and mummy in sows inseminated with different semen doses. The average number of born alive was 10.5, 11.0 and 10.4 from sows inseminated with 1.5×109, 2.0×109 and 2.5×109 sperms, re- spectively. Also, number of spermatozoa per dose did not affect litter size (p>0.10). There were no significant differ-ences of maternal genetic line difference on gestation period, total number born, number born alive, born dead and mummy. The estimated correlation coefficients of the different semen doses with total number born, number born alive, born dead and mummy were r=—0.00, —0.01, 0.02 and 0.02, respectively. Taken together, the result of this study suggested that when semen was appropriately inseminated after induced ovulation, insemination with low-dose (1.5∼2.0×109) semen dose not adversely affect sow’s fertility.