Urban tram wheel tyres operate under severe wear conditions resulting from complex wheel-rail interaction, particularly in modern low-floor vehicles with structurally constrained running gear systems. This study presents a probabilistic reliability-based framework for assessing wheel tyre durability as a function of running gear configuration, enabling statistically supported service-life estimation based on in-service data. Three representative low-floor tram designs (types A, B, and C) operating in a Polish urban network were analysed using a four-year dataset comprising 94 vehicles. Wheel profile degradation was evaluated through non-contact monitoring of key geometric wear indicators. Durability modelling was conducted within reliability theory using maximum likelihood estimation and statistical lifetime distribution fitting implemented in ReliaSoft Weibull++. Nominal durability at R(t) = 0.95 ranged from approximately 200,000 km to 319,000 km, representing a relative span of approximately 37% across configurations. Reduced durability was associated with limited yaw compliance in non-pivot bogies, whereas pivot-mounted designs improved self-centring and extended service life. Localised wear near suspended carbody sections further highlights the influence of load transfer mechanisms and vehicle structural configuration.