BACKGROUND During the SARS epidemic, healthcare utilization and medical services decreased significantly. However, the long-term impact of SARS on hospital performance needs to be further discussed. METHODS A municipal hospital in Taipei City was shut down for a month due to SARS and then became the designated SARS and infectious disease hospital for the city. This study collected the outpatient, inpatient and emergency service volumes for every year from April to March over four years. Average monthly service amount +/- standard deviation were used to compare patient volume for the whole hospital, as well as the outpatient numbers accessing different departments. The ARIMA model of outpatient volume in the pre-SARS year was developed. RESULTS The average monthly service volume of outpatient visits for the base year 2002 was 52317 +/- 4204 visits per month, and number for 2003 and the following two years were 55%, 82% and 84% of the base year respectively. The average emergency service volume was 4382 +/- 356 visits per month at the base year and this became 45%, 77% and 87% of the base year for the following three years respectively. Average inpatient service volume was 8520 +/- 909 inpatient days per month at the base year becoming 43%, 81% and 87% of the base year for the following three years respectively. Only the emergency service volume had recovered to the level of a non-significant difference at the second year after SARS. In addition, the departments of family medicine, metabolism and nephrology reached the 2002 patient number in 2003. The ARIMA (2,1,0) model was the most suitable for outpatient volume in pre-SARS year. The MAPE of the ARIMA (2,1,0) model for the pre-SARS year was 6.9%, and 43.2%, 10.6%, 6.2% for following 3 years. CONCLUSION This study demonstrates that if a hospital is completely shut down due to SARS or a similar disease, the impact is longer than previous reported and different departments may experience different recover periods. The findings of this study identify subspecialties that are particularly vulnerable in an infectious disease designated hospital and such hospitals need to consider which subspecialties should be included in their medical structure.