A 100-fold reduction of the standard dose for artificial insemination in pigs (3 x 10(9) spermatozoa in 80-100 ml fluid) can be used when spermatozoa are deposited surgically next to the uterotubal junction. The present study was performed to develop a technique for non-surgical deep intrauterine insemination in pigs without sedation of the animal. In Expt 1, sows were weaned, treated to induce oestrus and used to evaluate the difficulties involved in the insertion of a flexible fibre optic endoscope through the cervix and along the uterine horn. Deep uterine catheterizations were performed on each sow at 30-40 h after hCG treatment in the crate in which the animal was housed. The endoscope was inserted through an artificial insemination spirette, moved through the cervical canal and propelled forward along one uterine horn until the entire endoscope was inserted. In 30 sows (90.9%) no or minor difficulties were observed during insertion and in these animals the procedure was completed in 4.1 +/- 0.26 min. Insertion of the endoscope through the cervical canal was not possible in only one sow (3.03%). In Expt 2, endoscopic deep intrauterine insemination at 36 h after hCG treatment was performed in 15, 18 and 13 sows with 100, 20 or 5 x 10(7) spermatozoa, respectively, resulting in farrowing rates of 86.6%, 88.9% and 92.3%, respectively; there were no significant differences among groups. Farrowing rates after deep intrauterine inseminations were also not different from those achieved after standard intracervical insemination with 3 x 10(9) spermatozoa (control group: n = 48; 87.5%). Mean litter size (9.41 +/- 0.38 to 10.02 +/- 0.25) was also similar among the different experimental and control groups. In conclusion, endoscopic non-surgical deep intrauterine inseminations can be performed quickly in sows, and normal farrowing rates and litter sizes can be obtained after insemination with a small number of spermatozoa.