Using the empirical road network data and regression analysis, we investigated the ratio of the road distance to the straight line distance in major cities in Japan. The ratios are relatively different by cities. Therefore, in order to replace the road distance by the straight line distance, we do not use the known ratio that is revealed in different city without careful consideration. If we replace the road distance by the straight line distance, we have to consider the road network of the target city. In addition, we found that the cities which have small ratio of the road distance to the straight line distance generally tend to have high coefficient of determination. In other hands, the cities that have high ratio of the road distance to the straight line distance generally tend to have small coefficient of determination. Therefore, when we replace the road distance by the straight line distance in the cities which have small ratio of the road distance to the straight line distance, there is a certain degree of reliability.
