Groundbreaking technological innovations for train distancing and localization have the potential to radically boost dependability of Train Control Systems (TCSs), while actually posing notable challenges on dependability evaluation due to the uncertainty introduced on TCS vital parameters such as train position and speed. This paper focuses on applications of stochastic modeling and analysis for dependability evaluation of TCSs in presence of uncertainty, summarizing four representative case studies developed by the authors in previous publications. Without intent of completeness, this paper aims at showing the potential of quantitative evaluation methods in: i) representing and analyzing the variability of TCS parameters; ii) determining appropriate trade-offs between contrasting but equally relevant dependability-related attributes; iii) providing added value in synthesizing stochastic parameters from observed data; and, iv) supporting the definition of compositional solution methods for the analysis of complex TCSs. Along with a fairly extensive literature, this paper demonstrates the relevance and topicality of dependability evaluation of TCSs with uncertainty on vital parameters.