Seventy two lactating Manchega ewes were used to study the relationship between milk produced and total milk obtained with oxytocin. Ewes were maintained with their lamb suckling during weeks 1 and 2 postpartum (Period S). Between weeks 3 and 6, ewes were milked twice a day and lambs were allowed to suck after milking (Period SM). Lambs were weaned at week 6 and ewes were machine milked twice a day for the remainder of the lactation (Period M). Machine milked milk production was recorded weekly during Periods SM and M. Total milk production potential was measured after oxytocin injection in weeks 1, 2, 4, 6, 7, 9 and 11 of lactation. During Period SM, milk consumed by lambs was estimated by the double weighing method. Total oxytocin milk production decreased after partial (week 2) and complete (week 6) weaning. The drop in oxytocin milk production was larger during the complete versus the partial weaning (26 and 14%, respectively). The overall loss in oxytocin milk production between Period S and complete weaning (Period M) was 41%. The coefficient of correlation between milli and total milk production was high (r=0.48-0.91, P<0.001), being correlations obtained during the first 6 weeks of lactation. Milk production curve persistence of machine milked and oxytocin milk tended to be more similar as lactation advanced, which explained the improvement of the coefficient of correlation between machine milked and oxytocin milk production during the last part of lactation. Oxytocin milk was able to predict total milk produced within a wide range of production (R-2=0.80, P<0.001), but total milk production was consistently underestimated at levels of production of 1-2 L/d. Theoretical differences between milk produced and oxytocin milk increased an average of 94ml/d as the amount of oxytocin milk increased.