Deep intrauterine insemination (DUI) offers a suitable alternative for the commercial use of frozen-thawed boar semen. The present study evaluated how the ovarian status at DUIs of frozen-thawed spermatozoa (1 x 10(9) spz/dose, two DUIs, 30-31 and 36-37 h after detection of oestrus) in 179 sows would explain differences in fertility between two farms with similar, but not equal, reproductive management (experiment 1). A further experiment investigated whether an increase in sperm number per AI-dose (1 versus 2 x 10(9) spz/dose, two DUIs, 30-31 and 36-37 h after detection of oestrus, on 228 sows) could minimize this effect (experiment 2). Ovaries were checked by transrectal ultrasonography at the time of DUI and sows were classified into three categories: F-: ovarian pre-ovulatory follicles were visible during two examinations; O-: ovulation visible during one examination; and C-sows: corpora hemorragica visible during the two examinations. Overall farrowing rates differed (P < 0.01) between farms (70.1 versus 51.2%, farms A and B, respectively). Distribution of sows among ultrasonography categories also differed (P < 0.05) between farms (17.5, 72.2 and 10.3% were classified as F-, O- and C-sows in farm A, versus 40.2, 29.3 and 30.5% in farm B). Nevertheless, farrowing rates and litter sizes within categories did not vary between farms (P > 0.05). In addition, a two-fold increase in the number of spermatozoa per DUI improved (P < 0.05) fertility in F- and C-sows, but not in O-sows. In conclusion, the interval DUI-to-ovulation provides a major explanation for fertility differences between farms when frozen-thawed spermatozoa are used.