Study Objectives: We conducted a 5-year time study analysis of emergency department patient care efficiency. Our specific aims were (1) to calculate the main ED patient care time intervals to identify areas of inefficiency, (2) to measure the effect of ED and inpatient bed availability on patient flow, (3) to quantitatively assess the effects of administrative interventions aimed at improving efficiency, and (4) to evaluate the relationship between waiting times to see a physician and the number of patients who leave without being seen (LWBS) by a physician., Methods: Seven 1-week ED patient flow time studies were conducted from September 1993 to July 1998 using identical study design and methodology. Patients presenting with complaints of chest pain, abdominal pain, vaginal bleeding, and extremity injury were included to represent the level of severity of patient conditions seen in our Los Angeles County hospital ED. The calculated time intervals representing the main phases of evaluation and treatment were (1) triage presentation to completion of registration, (2) completion of registration to ED treatment area entry, (3) ED treatment area entry to initial medical assessment, (4) triage presentation to initial medical assessment, (5) initial medical assessment to disposition order, and (6) disposition order to patient discharge from the ED. Total ED lengths of stay (LOS) were also calculated as overall measures of efficiency. Time intervals were compared depending on the availability of ED and hospital inpatient beds. The effects of administrative interventions on the specific time intervals were assessed. The relationship between the median waiting time to see a physician and the number of LWBS patients was evaluated. Administrative interventions were implemented by a special interdepartmental continuous quality improvement committee. Interventions were aimed at specific sources of delay and inefficiency identified by the time studies., Results: Eight hundred twenty-six patients were included in the 7 time studies. The unavailability of ED and inpatient beds was associated with significant delays. There was a significant reduction of the median total ED LOS from 6.8 hours to 4.6 hours over the first 5 periods, presumably resulting from the administrative interventions. Median total ED LOS, however, increased from 4.6 hours to 6.0 hours during the last 2 periods, possibly as a result of an increase in our ED patient census and reductions in both nursing and physician staffing imposed by the recent Los Angeles County fiscal crisis. The number of LWBS patients was closely correlated to waiting time to see a physician ( r =0.79, beta=5.20, P =.033)., Conclusion: Time studies are an effective method of identifying areas of patient care delay. In our ED, targeted administrative interventions apparently reduced the total ED LOS and improved overall efficiency. Despite initial decreases in ED LOS, efficiency appeared to be adversely affected by reductions in nursing and physician staffing and increases in our patient census. The strength of the relationship between waiting times to see a physician and the number of LWBS patients suggests that decreasing waiting times may reduce the number of LWBS patients.