This study aims to explain the factors affecting ridership changes on the relatively new Taiwan High Speed Rail system (THSR). The analysis is based on monthly ridership data from January 2007 to December 2013. We also discuss the impact of THSR on competing modes such as air demand. Econometric time-series models are used for ridership estimation. Firstly a seasonal autoregressive integrated moving average (SARIMA) model was applied; showing that the ridership thrives and that the trend prediction fairly well performed if applied to data after 2012. Secondly, to specify the impact of explanatory variables, a first order moving average model was fitted. Results show that ridership, population and fuel price have a positive effect while unemployment and car ownership tend to reduce the THSR ridership. We include as separate factor "month since operation start", show that this factor is significant and discuss its relation to demand adaptation. Implications for general equilibrium modelling for new transport systems are discussed. Moreover, ridership data from two specific stations are used to test the importance of predominant trip purposes for demand estimation.