The search for small molecules that bind to the target is begun by screening libraries of potential drug compounds. This may be done by using the screening assay (a «wet screen»). In addition, if the structure of the target is available, a virtual screen may be performed of candidate drugs. Ideally the candidate drug compounds should be «drug-like», that is they should possess properties that are predicted to lead to oral bioavailability, adequate chemical and metabolic stability, and minimal toxic effects. Several methods are available to estimate druglikeness such as Lipinski’s Rule of Five and a range of scoring methods such as Lipophilic efficiency. Several methods for predicting drug metabolism have been proposed in the scientific literature, and a recent example is SPORCalc. Due to the complexity of the drug design process, two terms of interest are still serendipity and bounded rationality. Those challenges are caused by the large chemical space describing potential new drugs without side-effects.

Chills or cold flashes with goose bumps («cold turkey») alternating with flushing (hot flashes), kicking movements of the legs («kicking the habit») and excessive sweating are also characteristic symptoms. Severe pains in the bones and muscles of the back and extremities occur, as do muscle spasms. At any point during this process, a suitable narcotic can be administered that will dramatically reverse the withdrawal symptoms. Major withdrawal symptoms peak between 48 and 96 hours after the last dose and subside after about 8 to 12 days. Sudden withdrawal by heavily dependent users who are in poor health is very rarely fatal. Morphine withdrawal is considered less dangerous than alcohol, barbiturate, or benzodiazepine withdrawal.

It is difficult to study the performance effects of morphine without considering why a person is taking morphine. Opioid-naive subjects are volunteers in a pain-free state. However, most chronic-users of morphine use it to manage pain. Pain is a stressor and so it can confound performance results, especially on tests that require a large degree of concentration. Pain is also variable, and will vary over time and from person to person. It is unclear to what extent the stress of pain may cause impairments, and it is also unclear whether morphine is potentiating or attenuating these impairments.