Railway infrastructure is a complex system which comprises different subsystems. Long life span is one of the important aspects of this prime mode of transport. However, the useful life of its assets is highly dependent on the maintenance and renewal strategy used during the assets’ life cycle. Today’s demands on the railway industry call for increased capacity, including more trains, travelling at higher speeds with higher axle loads. This increased usage results in higher degradation of railway assets and higher maintenance costs. Formerly, railway maintenance procedures were usually planned based on the knowledge and experience of the infrastructure owner. The main goal was to provide a high level of safety, and there was little concern for economic issues. Today, however, the deregulated competitive environment and budget limitations are forcing railway infrastructures to move from safety limits to cost-effective maintenance limits to optimise operation and maintenance procedures. The goal is to make operation and maintenance cost-effective while still meeting high safety standards.One of the main parameters to assure railway safety and comfortable railway service is to maintain high quality of track geometry. Poor quality of track geometry, directly or indirectly, may result in safety problems, speed reduction, traffic disruption, greater maintenance cost and higher degradation rate of the other railway components (e.g. rails, wheels, switches and crossings etc.). The aim of this study is to develop a methodology to optimise track geometry maintenance by specifying cost-effective maintenance limits. The methodology is based on reliability and cost analysis and supports the maintenance decision-making process. The thesis presents a state-of-the-art review of track geometry degradation and maintenance optimisation models. It also includes a case study carried out on the iron ore line in the north of Sweden to analyse the track geometry degradation and discuss possible reasons for the distribution of failures along the track over a year. It describes Trafikverket’s (Swedish Transport Administration) maintenance strategy regarding measuring, reporting on and improving track quality, and it evaluates the efficiency of this strategy. It introduces two new approaches to analyse the geometrical degradation of turnouts due to dynamic forces generated from train traffic. In the first approach, the recorded measurements are adjusted at crossing point and then the relative geometrical degradation of turnouts is evaluated by using two defined parameters, the absolute residual area (ARa) and the maximum settlement (Smax). In the second approach, various geometry parameters are defined to estimate the degradation in each measurement separately. It also discusses optimisation of the track geometry inspection interval with a view to minimising the total ballast maintenance costs per unit traffic load. The proposed model considers inspection time and the maintenance-planning horizon time after inspection and takes into account the costs associated with inspection, tamping and risk of accidents due to poor track quality. Finally, it proposes a cost model to identify the cost-effective maintenance limit for track geometry maintenance. The model considers the actual longitudinal level degradation rates of different track sections as a function of million gross tonnes (MGT) / time and the observed maintenance efficiency.