In his initial formulation of the general theory of relativity, Einstein's proposal that freely falling gravitating massive bodies follow geodesic paths was submitted as an independent fundamental principle. By adopting this geodesic principle to supply the theory's law of motion, Einstein was immediately able to recover both the free-fall motion of bodies in non-relativistic regimes and the previously anomalous precession of the perihelion of Mercury. Over the last century numerous ostensible proofs claiming to have derived the geodesic principle from Einstein's eld equations have been developed. As a result physicists and philosophers of science alike frequently herald Einstein's theory for having the unique distinction of being able to derive its dynamical law of motion from its own eld equations.In this paper I critically survey the multiple attempts to derive the geodesic principle in the context of Einstein's theory. Grouping these results into three major families, which I refer to as (1) limit operation proofs, (2) 0 th -order proofs, and (3) singularity proofs, I argue that none of these strategies successfully demonstrates the geodesic principle, canonically interpreted as a dynamical law that massive bodies must actually follow geodesic paths in Einstein's theory.Speci cally, I argue for the following three claims: First, limit operation proofs fail to demonstrate that massive bodies are ever guaranteed to follow geodesic paths. Second, on the contrary 0 thorder proofs demonstrate that extended massive bodies generically deviate from uniformly geodesic paths. Moreover, the only potentially extended distributions of matter and energy that fail to avoid a uniform geodesic evolution are highly unstable, deviating from such motion under arbitrary perturbations of their angular momentum (or higher order moments). Third, thanks to certain mathematical theorems concerning distribution theory, alternative representations of massive bodies as unextended point particles must result either in precluding the possibility of coupling the particle to the spacetime metric in a way that is coherent with Einstein's eld equations or in having to excise the particle (and its would-be path) from spacetime entirely. This three pronged argument reveals that not only does the geodesic law of motion fail to be a deductive consequence of the eld equations, but also any attempt to canonically interpret the geodesic principle in such a way requires * In the following, M will be taken to be a smooth, orientable, four-dimensional manifold, and (M, g ab ) will be referred to as a Lorentzian spacetime if g ab is a smooth metric of signature (+, −, −, −) de ned on M. Excepting quoted material all further notational conventions follow that of (Wald, 1984).1 Forthcoming: Studies in History and Philosophy of Modern Physics that either the gravitating body is not massive, its existence violates Einstein's eld equations, or it does not exist within the spacetime manifold at all (let alone along a geodesic).Having rejected the canonical in...