This study was a retrospective analysis of drugs present in blood and urine samples taken from patients (n = 200) admitted to the emergency department of a major teaching hospital with a provisional diagnosis of deliberate self-harm. The aim was to assess the current limited drug screening strategy to see whether it needed to be changed in any way. Drugs present in blood and urine were identified by immunoassay or chromatography, categorized, and concentration-toxicity effects evaluated when practicable. For each case, the various drugs/drug classes detected were correlated with those reported by the patient. A questionnaire evaluation of doctor's perceptions of the influence of the primary blood drug screen on patient destinations was administered. The rapid primary drug screen using a blood/plasma sample detected some 46% of all drugs identified. The doctors considered that it was influential in deciding on immediate patient destination, and therefore, it is likely to be a cost-effective measure. In addition, the screen detected toxic concentrations of drugs in a significant proportion of patients who did not report their ingestion correctly. A primary drug screen using a urine sample detected opiates, cannabinoids, and amphetamines but such detection was considered unlikely to alter short-term treatment. A high-performance liquid chromatography and gas chromatography-mass spectroscopy secondary screen using blood and urine detected a significant number of additional drugs, but was slow, costly, and not likely to alter short-term treatment. The authors conclude that the primary screen for alcohol, benzodiazepines, paracetamol, salicylate, and tricyclic antidepressants remains the optimal drug screening strategy. Quantitative or qualitative estimation of patient-reported drugs such as quinine, theophylline, verapamil, and antiepileptics may be justifiable in individual patients.