Milk samples from all quarters of 13 dairy cows, each treated with an antibiotic preparation for clinical mastitis in one quarter, were collected at each milking during treatment and for at least six milkings thereafter. After storage at —25 °C, these were tested for the presence of antibiotics. Transfer to untreated quarters was found in eight of the 13 cases studied, although in only two cases was the level in untreated quarters greater than 0.01 μ/ml. The main factor affecting transfer was found to be a history of infection in the treated quarter. This was observed in all but one case where transfer had occurred. During the period of study incorporation of milk from untreated quarters would not have raised the antibiotic level in the bulk supply to a level which would have incurred a financial pena