Clinical tests of gastrointestinal permeability that rely on the urinary recovery of enterally administered probes can yield invalid results in critically ill patients.