The degradation processes affecting railway track condition depends both on the resistance of the track and on the stresses subjected to it. Regarding the stresses, both their magnitudes and cycles are of importance when considering the degradation. Furthermore, the stresses have some regularity and variability in the time domain, while the degradation resistance of a track has some spatial regularity as well as variability. In addition, the condition measurements of track may be both irregular and contain measurement errors. Hence, it is challenging to model the condition of track to enable predictions and condition-based maintenance. However, wear prediction models could help to change large parts of the maintenance practice from predominantly corrective to preventive if both the deterministic and the stochastic components of the wear process can be estimated with sufficient accuracy. In this study, one-step-ahead predictions have been used for establishing prognostic models based on repeated measurements of railway track geometry to estimate track wear properties, degradation rates and stochastic behaviour including measurement errors. The prognostic models have then been used for condition assessment and state predictions. Repeated sampling allows for estimations of measurement errors, but the irregular sampling need to be accounted for by interpolation in the time series modelling approach