INTRODUCTION    In local cities in Japan, it is urgent to make public services more efficient in order to respond to the declining birthrate and aging population as the declining population. Demand for emergency transportation, which is one of public services, has been increasing in recent years. From this background, the purpose of this study is analyzing the optimum location of fire departments. The point of view of this research is to minimize the response time by confirming emergency services dispatched and absent situation.<br> METHODS    We calculated the average response time for the optimization algorithm with dispatch simulation based on past actual dispatch data. The step outline of the dispatch simulation is as follows.<br> 1. Confirm emergency services dispatched and absent situation.<br> 2. Demand occurs.<br> 3. Calculate traveling time from all emergency services to the demand place<br> 4. Search for emergency services with the nearest.<br> 5. Dispatch the nearest emergency services, if they can dispatch.<br> 6. If they cannot dispatch, Search for the 2nd nearest emergency services.<br> 7. If they cannot dispatch, Search for the next one.<br> 8. Repeat steps 1 to 7 to end of the year.<br> In addition, the unit of evaluation was taken as the average response time for one year. Furthermore, we compared with the p-median model which is a conventional location model. It is possible to reproduce the place of demand generation and the allocation of the ambulance in this simulation. This makes it possible to optimize the extending distance of travel distance compared with the nearest ambulance.<br> RESULTS    The results are summarized below.<br> Average travel distance of the year / Maximum travel distance of the year<br> Current location (13 facilities and 13 units): 2798.11 m / 20343.96 m<br> Median model (13 facilities and 13 units): 2431.31 m / 19944.68 m<br> Simulation model (13 facilities and 13 units): 2360.12 m / 18750.97 m<br> Simulation model (12 facilities and 13 units): 2398.93 m / 18750.97 m<br> Median model (12 facilities and 12 units): 2604.14 m / 22536.70 m<br> Simulation model (12 facilities and 12 units): 2506.33 m / 20405.53 m<br> The results show the median model shortened distances is about 367 m and the simulation model shortened distances is about 438 m compared with the current location. It can shorten distances about 399 m compared with the current location even if one of those facilities is reduced and two ambulances are arranged at a facility which one of the other. Furthermore, it was found that it is possible to shorten about 292 m even in the case of abolishing one of those facilities and one of those ambulances.<br> CONCLUSION    We indicated that there were many cases where demand occurred during dispatch, and the delay of arrival time to the site due to this. From this, it was suggested that it is important to consider not only the nearest facility location but also the next facility location when we decide on the location of the fire department for reducing the travel distance. In addition, it can shorten the arrival time to the demand place by preventing cases that cannot dispatch during the dispatch to the other demand. In other words, not only the location of the facility but also adding the ambulance have an effect on shortening the arrival time.