Assigning functions to every gene in a living organism is the next challenge for functional genomics. In fact, 85-90% of the 19,000 genes of the nematode Caenorhabditis elegans genome do not produce any visible phenotype when inactivated, which hampers determining their function, especially when they do not belong to previously characterized gene families. We used 1 H high-resolution magic angle spinning NMR spectroscopy ( 1 H HRMAS-NMR) to reveal the latent phenotype associated to superoxide dismutase (sod-1) and catalase (ctl-1) C. elegans mutations, both involved in the elimination of radical oxidative species. These two silent mutations are significantly discriminated from the wild-type strain and from each other. We identify a metabotype significantly associated with these mutations involving a general reduction of fatty acyl resonances from triglycerides, unsaturated lipids being known targets of free radicals. This work opens up perspectives for the use of 1 H HRMAS-NMR as a molecular phenotyping device for model organisms. Because it is amenable to high throughput and is shown to be highly informative, this approach may rapidly lead to a functional and integrated metabonomic mapping of the C. elegans genome at the systems biology level.functional genomics ͉ metabolic profiling ͉ metabonomics ͉ nuclear magnetic resonance A ssigning functions to every gene in a living organism is the next challenge for functional genomics. However, only a small proportion of genes produce visible phenotypes when inactivated; for example, only 10-15% of the 19,000 genes of the nematode Caenorhabditis elegans produce a visible phenotype (1, 2). For the remaining ones, determining their function is more difficult, especially when they do not belong to previously characterized gene families.For example, oxidative stress is a key, yet subtle, biological process involved in aging, with long-term integration of many slow processes leading to irreversible cellular and molecular damage.Phenotyping plays a critical role in postgenomic sciences. Today, a range of molecular phenotyping tools is available to characterize mutations. Although gene expression and protein profiling are predominantly used (3-5), metabonomic and metabolomic strategies (6, 7) advantageously produce metabolic fingerprints that allow identification of variations in low-molecular-weight compounds in biofluids or organs in response to pathophysiological events (8), drug treatments (9), or genetic polymorphisms (10). It is therefore an attractive hypothesis-free approach for large-scale functional genomics in model organisms (11).Here we capitalize on recent developments in solid-state NMR that allow the acquisition of highly resolved 1 H spectra of metabolites from inhomogeneous materials such as biopsies (12) or food (13). As shown below, when applying 1 H high-resolution magic angle spinning NMR spectroscopy ( 1 H HRMAS-NMR), we provide complex metabolic phenotypes (6) or metabotypes (8) suitable for discriminating between C. elegans oxidative stress mutants, w...