Determining the probability that a new drug will be abused is an important step in reducing the overall abuse of therapeutic drugs. Since the likelihood that a drug will be abused by a patient must be carefully weighed against the benefit provided by the drug, it is important that research outline any and all reinforcing effects a drug may have which could lead to subsequent abuse. Prediction of the likelihood of abuse has historically been based upon human experiments and observation. This method, however, is increasingly being replaced with experimentation on animals.

Research conducted since the early 1960s has shown that animals such as monkeys and rats will, with very few exceptions, repeatedly self-administer the same drugs that human beings are likely to abuse. Moreover, test animals do not self-administer drugs that human beings...