This study aimed at quantifying the variation in test-day milk yield (TDMY) of dairy cows according to Map-infection status of cows. The cow-status was determined combining (i) her testing(s)-result(s) (ELISA, faecal culture (FC), PCR, Ziehl staining), (ii) the Map-status of her herd, and (iii) her possible vaccination against Map. When combining this information, a total of 14 cow-statuses were defined. A total of 23,219 cows in 569 herds located in Bretagne was considered. The effect on TDMY of the cow-status to Map, adjusted for herd-year (random), lactation number, days in milk and breed was assessed using mixed linear models. The average TDMY was significantly lower in cows from herds with at least one Map-infected cow (defined as positive herds). Individual TDMY showed a reduction of 1.6, 2.5, 2.1 and 6.2 kg/day (P<0.001) for cows tested negative in a positive herd, not vaccinated and ELISA-positive, PCR- or FC-positive, and Ziehl-positive, respectively, in comparison with cows in Map-negative herds. ELISA-tested positive but vaccinated cows had a smaller loss in TDMY than those not vaccinated, suggesting the coexistence of infected and not-infected animals among ELISA-positive cows. These estimates will be used to further assess the economic impact associated with Map-infection in dairy herds or to enlighten the culling decision making process regarding infected cows.