Previous studies have either failed to allow for such factors or have failed to compare opioid users to adequate control groups. From interviews with 151 Scottish prisoners and non-prisoners, divided into five drug-using groups: (1) alcohol only; (2) cannabis and alcohol; (3) other drugs but not opioids; (4) moderate opioids; (5) heavy opioids, data were obtained on drug use frequency, crime frequency, and a variety of other variables. It was found that: (1) there were no differences between incarcerated and non-incarcerated opioid users, in fact incarceration had no substantial effect on other variables; (2) heavy opioid users committed crimes significantly more frequently than did moderate opioid users, non-opioid polydrug users, cannabis users, or alcohol users. But, moderate opioid users did not commit crimes significantly more frequently than did the other groups; (3) drugs were related to crime. Polydrug use (including cannabis use) was more related to theft and delinquency than was opioid use. Alcohol use was related to fraud; (4) there were larger explanations of crime than opioid use -- criminal and polydrug use -- and crime was a better explanation of opioid use variance than opioid use was of crime. It is concluded that the need for opioids did not simply cause crime. Rather, crime and opioid use tended to influence each other. However, this relationship was not special to opioids but, depending on historical circumstances, could -- and to some extent does -- apply to any drug. In consequence, society's treatment of drug-using criminals needs to deal with drug use and crime together. 2 tables, 41 references. (Author abstract)