Although much research on the drugs-crime link does not permit generalizations due to methodological limitations, a number of generalizations are apparently empirically based. One finding is that users of illegal drugs consist of persons with widely varying habit sizes and frequencies. A significant percentage of the income required to support large drug habits is obtained from selling drugs. Although the amount of crime attributed directly to drug use has been overstated, drug users are significantly involved in crime. In many cases, this is a continuation of criminality that existed prior to drug use, but levels of criminal activity correlate with levels of drug use. There is no consistent evidence to link violence with the direct effects of illegal drugs, with the possible exceptions of barbiturates and amphetamines. Findings support the conclusion that under a system of drug use control that criminalizes drug trafficking, addicts must commit income-producing crime to pay for their drugs. This suggests that a criminalization policy should be reexamined based on a realistic view of how legalization will impact crime. The possibility of having differing control policies for various drugs should be considered. There is a need for Australian drug policy to be based on longitudinal and ethnographic studies of Australian drug users selected from unofficial sources. 1 table and 38 references.