Five lactating goats which had kidded normally in March were mated during seasonal anoestrus in May, at the time of peak milk production, after ovulation had been induced using gonadotrophin-releasing hormone (Knight et al. 1988). Milk yield was unaffected by the hormone treatment, and decreased at the same rate as that of control (non-pregnant) goats for the first 8 weeks of the pregnancy. Thereafter yield declined more quickly in the test goats and just before parturition (in October) was 57% of the control value. Following parturition in the test animals, yield rose rapidly as the second lactation was established. None became 'dry' at any stage. Yield continued to decline with advancing lactation in the controls, which were mated normally in October or November and dried-off in December. During their second ('extra') lactation in the winter the test animals produced 12% less than in a normal second lactation in summertime; during the year the extra lactation meant that the test animals produced 73% more milk than the controls. In some, a second concurrent pregnancy was established during the extra lactation, with the result that three lactations were obtained in the time normally taken for two. Mammary cell number and proliferation rate were both higher in the pregnant animals than in the controls in week 23 of the first lactation.