In organic dairy farms, cow-calf contact is encouraged until weaning and requested by society. However, farmers question this practice, especially because of the loss of marketable milk. At INRAE experimental farm ‘Herbipole’, we tested two different suckling strategies on animal performance and behaviour. A 14-cow ‘Classic’ rearing system (C group) was compared during 14 weeks to two suckling systems. In the C group, calves were separated from dams immediately after birth and fed with an automatic milk feeder until weaning. In the ‘Dam’ group (D), dam-calf contact was allowed from birth to weaning, between morning and evening milking. In the ‘Mixed’ group (M), calves were kept with dams until three weeks (as in D group) before being separated and reared as in C group. All calves were weaned at about 11 weeks. On average, over 14 weeks, D and M cows produced 25.1% less milk at parlour than C cows; milk fat content was 3.6 g/kg lower in D group compared to C and M, and milk protein content was intermediate between C and M. After 11 weeks, D-calves weighed 20.5 kg more than C and M calves. Cows and calves both vocalised for one week after separation or after weaning. All calf vocalisations were at a maximum during the first four days. Cows’ vocalisations were less notable when calves were removed after three weeks compared to 11. In conclusion, a three-week suckling period seems better for farmers’ income and cows’ distress, but it induces stress for calves at both separation and weaning, without benefit on growth.