The Michaelis-Menten (MM) equation is the basic equation of enzyme kinetics; it is also a basic building block of many models of biological systems. We build a stochastic and microscopic model of enzyme kinetics inside a small subcellular compartment. Using both theory and simulations, we show that intrinsic noise induces a breakdown of the MM equation even if steady-state metabolic conditions are enforced. In particular, we show that (i) given a reaction velocity, deterministic rate equations can severely underestimate steady-state intracellular substrate concentrations and (ii) different reaction schemes which on a macroscopic level are indistinguishable because they are described by the same MM equation obey distinctly different equations in subcellular compartments.