OBJECTIVE: Adequate treatment of patients' pain is a top priority for the World Health Organization (WHO), American Medical Association (AMA), and American College of Emergency Physicians (ACEP), but "adequate" is not clearly defined. Most previous studies of emergency department (ED) pain treatments have centered on musculoskeletal pain in terms of rates of analgesia and disparities in treatment based on race and age. This study will examine complaints of pain other than musculoskeletal and will focus on treatment disparities that may result from differences inpatient and physician characteristics. METHODS: This retrospective study is of ED patients 18 years and older with nonmusculoskeletal pain who were seen by ED faculty over a period of eight weeks. Logistic regression and CHI2 tests were performed to quantify effects of doctor, patient, and clinical characteristics on rates of ED analgesia, ED opioids, and analgesic prescriptions at discharge. RESULTS: A total of 1360 patients were included. There was wide variation in the type and frequency of ED analgesia depending on the attending doctor. For example, patients seen by one specific ED doctor were less than half as likely to receive any analgesia and seven times less likely to receive an opioid than those seen by another doctor. Age, race, doctor's training and experience, and whether the patient had chronic pain were important predictors of ED analgesia. There were similar findings for ED opioids and discharge analgesics. CONCLUSION: Pain practices in EDs are highly variable and seem inadequate when measured against the goals of WHO, AMA, and ACEP. Patient age, race, and type of pain and the physician's identity, training, and experience all contribute to practice variation. Further research is needed to identify the causes of these variations, and there is a need to develop interventions to standardize and improve pain assessment and treatment.