Two experiments were conducted to study the amount of aflatoxin M1 (AFM1) in milk in response to feeding aflatoxin B1 (AFB1). In experiment 1, four dairy ewes in early lactation received a single dose of pure AFB1 (2 mg). Individual milk samples were collected during the following 5 d to measure AFM1 concentration. The average excretion of AFM1 in milk followed an exponential decreasing pattern, with two intermediate peaks at 24 and 48 h. No AFM1 was detected in milk at 96 h after dosing. The mean rate of transfer of AFB1 into AFM1 in milk was 0.032%, with a high individual variability (SD = 0.017%). In experiment 2, 16 dairy ewes in midlactation were divided into four groups that received different daily doses of AFB1 (0, 32, 64, and 128 microgram for control and groups T1, T2, and T3, respectively) for 14 d. Pure AFB1 was administered to each animal divided in two daily doses. Individual milk samples were collected at 12, 24, 36, 48, 72, 96, 144, 216, and 312 h after the first AFB1 administration, during the intoxication period, and every 24 h for 7 d after the withdrawal of AFB1. AFM1 was detected in the milk of all animals of the treated groups at 12 h after the administration of AFB1. In all treated groups, milk AFM1 concentration increased from 12 to 144 h after the beginning of administration. It then decreased, reaching a stable concentration at 216 and 312 h after the first administration. No AFM1 was detected in milk 3 d after the last administration of AFB1. Milk AFM1 concentration measured at steady-state condition was significantly affected by the AFB1 dose (0.031, 0.095, and 0.166 in T1, T2, and T3 groups, respectively), with a linear relationship between AFB1 dose and milk AFM1 concentration (R2 = 77.2%). The carryover (AFM1/AFB1 ratio) was not significantly affected by treatment, and its mean value was 0.112% (SE = 0.011). The carryover was lower than that reported for dairy cattle and goats, suggesting a better ability of sheep to degrade AFB1.