Substrate profiles for enzymes are typically determined in vitro. However, the extent to which these data correlate with substrate utilization in vivo, where many competitive metabolic pathways operate in parallel, remains unknown. Here, we describe a standard-free metabolite profiling strategy to identify endogenous substrates of enzymes and its application to discover several brain lipids, both known and novel in structure, that are catabolized by fatty acid amide hydrolase (FAAH) in vivo. Remarkably, no correlation was found between the rates at which FAAH hydrolyzed lipids in vitro and their likelihood to serve as FAAH substrates in vivo. These data indicate that an enzyme’s endogenous functions cannot be directly extrapolated from its properties in vitro and showcase how global metabolite profiling addresses this important challenge.