<p>The main purpose of this study is to clarify the actual fare level and factors affecting the fare level in Japan. In numerical calculations, Ramsey pricing is applied for different demand groups by using a database of 133 railways between 2015 and 2020. Ramsey pricing is intended to maximize economic welfare, under the constraint that fare revenues cover the costs of service provision. First, among three different demand groups, the fare levels are highest for non-commuter regular fares, followed by work commuter pass fares, and then student commuter pass fares. Another important finding is that the markup rate for student commuter pass fares is only about 50% of the marginal cost, while the non-commuter regular fares are about twice the marginal cost. This result might show the possibility of cross-subsidy between demand groups in fare setting. As for factors affecting the fare level, seemingly unrelated regression (SUR) analysis for fares of the three demand groups is applied by using 732 observations between 2015 and 2020. The SUR method is one regression method to estimate a system of several equations for these services. The results show that both, a longer network length and a higher share of rail transportation work increase fare levels. However, neither a higher profit level of railway companies nor public ownership has the effect of lowering fares, but in fact have the opposite effect of raising fares. Yardstick regulation works in the direction of reducing fares, as expected by the government, but the effects of the vertical structure organizational type are inconclusive.</p>