Light rail trains (LRT) are an important part of public transport but due to perceived high life-cycle costs are not always considered suitable. Life cycle cost reduction might be achieved through a knowledge-based maintenance management rather than just on experience. This work develops limits of maintenance and renewal of LRT systems based on vehicle reactions to the current track quality through measured data, multibody simulations and track geometry indices. An approach based on knowledge would lead to a track condition which allows a safe, comfortable, and under an appropriate maintenance strategy, economically profitable operation.