Two experiments were carried out to compare mechanical milking in mid-level (ML) and low-level (LL) milkline in goats. The first trial used 40 intramammary infection (IMI)-free goats that had been milked in ML during a pre-experimental period of 4±1 weeks post partum. These animals were divided into two groups (n=20), randomly assigning each group to ML or LL milking for a 17-week experimental period. During this period, several strategies were applied to increase teat exposure to pathogens in both experimental groups. The IMI rate was the same in both experimental groups (30% of goats), although the majority of new infections appeared earlier in ML (weeks 1-5) than in LL (weeks 7-16). Teat-end vacuum range (maximum minus minimum vacuum) was higher in ML than in LL, but no significant differences were found in the remaining variables [milk production and composition, somatic cell count (SCC), frequency of liner slips+teatcups fall off]. In the second experiment, in a crossover design (54 goats in fourth month of lactation; 2 treatments, ML and LL, in 2 experimental periods each lasting 1 week) it was observed that both the milk fractioning (reduced machine milk and increased machine stripping) and average machine milk flow worsened slightly in ML milking; in contrast, no differences were observed in total milking time or teat thickness changes after milking. It was concluded that the differences found between ML and LL are not sufficiently important to discourage breeders from using ML in goat milking.