Breast feeding has increased by approximately 25% in the United States during the past decade and this trend appears to be continuing. The number of drugs available to lactating women is also growing at a rapid pace. The excretion of drugs into breast-milk presents a potential danger to infants. In spite of this, little is known about the excretion of drugs into breast-milk. The ability to predict which drugs are potential hazards would be very useful in the clinical setting. This study quantitatively correlates the human milk to plasma concentration ratio of various basic and acidic drugs (log M/P) with the square root of the molecular weight, the partition coefficient (log P) and the degree of dissociation (log U/D). For basic drugs there is a negative-dependence on both log P and log U/D. High lipophilicity favours protein binding and reduces the amount of drug available for diffusion into milk. Therefore, as log P increases, the log M/P decreases. The negative-dependence on log U/D indicates that the higher the degree of dissociation of the base in plasma, the greater the log M/P will be. This fits well with the concept of ion-trapping. A strong base is more likely to be transferred and then trapped in milk which has a lower pH than plasma. For acidic drugs there is a negative-dependence on both square root (MW) and log P. The negative-dependence on square root (MW) suggests that large molecules are less likely to be able to diffuse into the milk. A negative-dependence on log P appears to hold true for bases and acids. Log M/P decreases as log P increases. This is probably due to increased protein binding by lipophilic drugs through non-specific hydrophobic interaction with plasma protein.