T his article reviews the information available on fetal behavioral effects of drugs that are subject to substance abuse in human pregnancy. It deals primarily with the effects that occur at the time of in utero drug exposure, and not with the longer term postnatal consequences of such exposure. Although it is the latter area that has received most attention, both in animal and human investigations,‘,* an understanding of the effects of drugs on fetal behavior is also important. Because most tests of fetal health monitor aspects of fetal behavior, drug-induced behavioral alterations could interfere with these tests and hinder the identification of fetuses at risk. Also, prolonged fetal exposure to drugs that affect the central nervous system (CNS) can lead to fetal tolerance/dependence, with the associated risk of abstinence symptoms when drug exposure decreases or ends. Before delivery, this can lead to fetal distress and increased perinatal mortality and morbidity.‘r4 Finally, prolonged drug-induced alterations in fetal behavior patterns may be one of the mechanisms for long-term neurologic dysfunction after birth. While many drugs prescribed to or taken by pregnant women for legitimate medical reasons can affect fetal behavior, the major focus of concern in this area relates to social or illicit drugs that are subject to abuse. The list includes among others ethanol, marijuana, heroin, methadone and other opiates, amphetamines, phencyclidine, and cocaine, with the latter drug currently being of most concerm5 However, there are data on fetal behavioral effects for only ethanol, morphine, methadone, and cocaine, and this review will concentrate on these. Also, because the impact of these drugs on the fetus depends in part on the extent of maternal/ fetal transfer, this subject will also be reviewed. Evidence that maternal substance abuse during pregnancy can have profound effects on the developing brain have come primarily from two sources: (1) experimental studies employing small animal models (primarily rats and mice), in which drugs are given to pregnant animals or to newborns, with subsequent assessment in the offspring of behavioral parameters, such as locomotion and learning ability;‘,‘j and (2) follow-up studies of human newborns from pregnancies associated with maternal substance abuse.‘.” However, neither of these approaches have provided data on drug-induced alterations in fetal behavior. For this, two other sources of information are relevant. There are studies of chronically instrumented, large animal species, primarily sheep, in which fetal behavioral parameters in utero and the effects of drugs on these parameters can be monitored. Maternal/fetal drug disposition and dose-response relationships can also be examined.‘0z” It is also possible to use different routes of drug administration; for example, fetal administration can be employed to assess the direct fetal behavioral effects of drugs, which when given to the mother, elicit maternal effects that can alter fetal behavior.‘* Different temporal patterns of drug administration can also be examined, to model different patterns of substance abuse in pregnant women, for example, chronic as opposed to single episode administration.‘3z’4 However, there are several caveats to the extrapolation of data obtained from pregnant animals to the human situation. The structural and functional characteristics of the epitheliochorial placenta of sheep and other ruminants are markedly different from those of the hemochorial placenta of humans, so that data on the extent of maternal/fetal transfer of drugs in sheep are not always applicable to humans.15 For lipophilic drugs this is less of a problem, and most drugs subject to substance abuse are in this