Opioid-induced immunosuppression has been demonstrated in cell culture experiments and in animal models. This is in striking contrast to the paucity of confirmatory studies in humans. This review describes the basic pharmacokinetics and -dynamics of opioid use in patients. It summarizes the major findings on opioid use and infectious complications in intensive care unit (ICU) patients, in patients with acute or chronic non-malignant pain, and in intravenous drug users (IDU). The limitations of studies in each area are discussed. For example, ethical concerns may complicate randomized placebo-controlled trials (RCT) in acute postoperative pain and for a large part of ICU patients. Importantly, most studies in patients with chronic (non-malignant) pain only inadequately report infectious complications in relation to opioid use since their incidence is usually not considered to be drug related. Infectious complications in IDUs are very frequent but cannot easily be distinguished from risk behavior or risk environment. In summary, convincing clinical evidence is lacking that opioids per se increase the rate of infectious complications in most patient categories. From a clinical standpoint, important unresolved issues are i) selection of relevant animal models, ii) opioid selection and discontinuation, and iii) the role of coexisting diseases and concomitant other medications.