Previous studies of the GTPase reaction catalyzed by p21ras have indicated that the logarithm of the observed reaction rate and the pKa of the bound GTP are correlated by the Brønsted relationship log(kcat) = beta pKa + A. While most of the Ras mutants display a Brønsted slope beta of 2.1, a small set of oncogenic mutants exhibit a beta of > > 1. On the other hand, it was found that the corresponding Brønsted slope for the GTPase reaction of p21ras in the presence of GTPase Activating Protein (GAP) is about beta = 4.9. The present work explores the basis for such linear free energy relationships (LFERs) in general and applies these concepts to p21ras and related systems. It is demonstrated that the optimal way to analyze LFER is by using Marcus type parabolas that represent the reactant, intermediate, and product state of the reaction in a relevant energy diagram. The observed LFER is used to analyze the actual free energy surface and reaction path of the intrinsic GTPase reaction in p21ras. From this, a model reaction profile can be constructed that explains how a LFER can arise and also how the different observed Brønsted coefficients can be rationalized. This analysis is augmented by solvent isotope effect studies. It is pointed out that the overall activation barrier reflects the energy of the proton transfer (PT) step, although this step does not include the actual transition state of the hydrolysis reaction. The proposed GTP as a base mechanism is compared to a recently proposed reaction scheme where Gln61 serves as a proton shuttle in a concerted mechanism. It is shown by unique energy considerations that the concerted mechanism is unlikely. Other alternative mechanisms are also considered, and their consistency with the observed LFER and other factors is discussed. Finally, we analyze the observed LFER for the GTPase reaction of p21ras in the presence of GAP and discuss its relevance for the mechanism of GAP activation.