Objective: For many patients with addiction and other substance problems, the emergency department (ED) is the sole provider of medical care. This study sought to determine the prevalence and characteristics of substance-related medical problems in ED patients, as defined by documentation in the medical record. We also sought to compare the ED resource use (length of ED stay and number of revisits) of patients with and without substance problems. Methods: Trained evaluators using explicit criteria reviewed all ED charts during a 6-week period at a Canadian tertiary care teaching centre. Data was collected on demographics, documentation of problematic substance use and whether the ED visit was due to substance problems. Using a computerized database, we determined how many patients with and without substance problems had 1 or more subsequent ED visits during the 1-year period from Sept. 1, 2002, to Aug. 31, 2003. Results: Of 6064 visits made by 5194 patients, 6026 visits (99.4%) representing 5188 patients (99.9%) were captured for review. Of those visits, 674 (11.2%, 95% confidence interval [CI] 10.4%-12.0%), made by 600 patients, had documentation of problematic substance use and 521 visits (8.6%, 95% CI 7.9%-9.4%) by 469 patients were caused by substance problems. The mean age of patients with a visit due to a substance problem was 39.2 years, compared with 48.5 years for those with other visits (p < 0.001). The admission rate for substance-related visits was 25.3%, compared with 17.6% for other visits (p < 0.001). For discharged patients, the median length of the ED visit owing to substance-related problems lasted 232 minutes (IQR [interquartile range] 267 min), compared with 164 minutes (IQR 167 min) for other visits (p < 0.001). In 1 year of follow-up, 161 of 600 patients (26.8%) with a substance problem made 466 revisits (mean 0.78 revisits/patient), compared with 975 of 4588 patients (21.3%) without a substance problem who made a total of 2150 revisits (mean 0.47 revisits/patient, p < 0.001). Conclusion: Substance problems contribute significantly to ED visits, hospital admissions and duration of ED stay at a tertiary centre. It is likely that our methodology underestimates the scope of the problem and that a universal screening program would find a higher prevalence. The magnitude of this problem supports the need for an interdisciplinary identification and intervention program for ED patients with substance-related issues.