Artificial insemination (AI) in sheep is currently limited by the poor fertility obtained following non-surgical intracervical insemination of frozen-thawed semen. An exception to this general finding is the non-return rate of around 58% reported for large scale on-farm AI in Norway. The objective of the present study was to determine if similar results could be obtained under Irish conditions. Comparisons were made between semen collected, and frozen, from rams in Norway (NOR) and Ireland (IRL). The effects of synchronisation and inseminator were also examined. Parous ewes (n=297) of various breed types were inseminated to a natural (N) or synchronised (S) oestrus with either fresh (from Irish rams) or frozen-thawed (IRL and NOR) semen. Ewes were randomly assigned, within breed, to the following treatment groups: (i) Fresh-N: n=28, (ii) Fresh-S: n=30, (iii) IRL-N: n=62, (iv) IRL-S: n=50, (v) NOR-N: n=68, (vi) NOR-S: n=59. Within each group, ewes were inseminated by an experienced Norwegian or by an Irish inseminator. Pregnancy rate did not differ significantly between ewes inseminated to a natural or synchronised oestrus nor between Norwegian and Irish frozen semen. The proportion of ewes pregnant after insemination with fresh semen was 0.82 and 0.70 (treatments i and ii) compared with 0.40, 0.52, 0.34 and 0.37 (treatments (iii)-(vi)) for frozen semen (P<0.001). Corresponding litter sizes (+/-S.E.), adjusted for ovulation rate, were 2.9+/-0.22, 3.3+/-0.23, 2.2+/-0.21, 1.7+/-0.21, 2.2+/-0.21 and 2.1+/-0.21 (fresh versus frozen; P<0.001). There was an interaction between semen type (fresh or frozen) and oestrus type (N or S) for litter size due to an increased adverse effect of frozen semen on litter size in synchronised ewes (P<0.05). Pregnancy rate was significantly influenced by breed of ewe (P<0.01) and inseminator (P<0.05). These results suggest that ewe breed may be a critical determinant of the potential for the exploitation of cervical insemination of frozen-thawed semen in sheep breeding programmes.