A hedonic pricing model is estimated based on sales data from three metropolitan areas in the Netherlands (Amsterdam, Rotterdam and Enschede) to analyse the effect of railway accessibility on house prices. Railway accessibility is measured by both the distance to a railway station and an index of quality of railway services provided at the station. Two railway station considerations were taken: the nearest railway station and the most frequently chosen railway station. Correcting for a wide range of other determinants, the model based on the most frequently chosen station outperforms the model based on the nearest railway station in estimating the effect of railway accessibility. The dissimilarity between the results of the two models increases with the increase in the urbanisation level of the metropolitan area. © 2010 Urban Studies Journal Limited.