Two artificial lighting regimens were studied in a commercial unit of 800 sows. The aim was to develop a simplified lighting regimen to overdrive the effects of season on reproduction. A long-day group had a constant 16 hours light and eight hours dark photoperiod in all units during a production cycle. A short-day group had eight hours of light and 16 hours darkness in a farrowing unit for four weeks and in a mating unit for four weeks. After one month of pregnancy the short-day group was transferred to 16 hours light and eight hours dark for the rest of the dry-sow period. Production data were collected for eight months, and the farrowing rate, weaning to oestrus interval, culling rate for fertility problems and the number of live-born piglets were analysed. Seasonal infertility, which had previously affected the herd, was not detected in either group during the follow-up period. The farrowing rate was 90 per cent for both groups. The median weaning to oestrus interval was five days in both lighting regimens with ranges from four to 74 days in the short-day group and three to 55 days in the long-day group. The long-day programme had a positive effect on the weaning to oestrus interval. When the sows were classified as either "normal", with a weaning to oestrus interval of up to 10 days, or "problem" animals, with a weaning to oestrus interval of over 10 days, the short-day programme also had a positive effect. The culling rate for fertility problems was 2.4 per cent for the short-day group and 3.2 per cent for the long-day group (P=0.027). The only significant predictor for the number of live-born piglets was parity (P=0.027).