The monitoring system of a rail vehicle and track has been developed under the project MONIT - Monitoring of Technical State of Construction and Evaluation of its Lifespan. The main subsystem of a rail vehicle being the object of on-line monitoring is the suspension. The principle of operation of the monitoring system is based on the statistical analysis of acceleration signals. The selection procedure of damage-sensitive parameters of signals is introduced. In order to improve suspension condition assessment, selected parameters are analysed in multidimensional space. As a result, a new diagnostic indicator is analysed - Euclidean metric between reference point (nominal state) and a point representing current measurement (condition). The proposed fault detection method was used for analysis of recorded data during an experiment on wagons with suspension damages.