We investigate the mutational structure of the Caenorhabditis elegans metabolome by means of a mutation accumulation experiment. We find that pool sizes of 29 metabolites vary greatly in their vulnerability to mutation, both in terms of the rate of accumulation of genetic variance (the mutational variance, VM) and the rate of change of the trait mean (the mutational bias, ΔM). Strikingly, some metabolites are much more vulnerable to mutation than any other trait previously studied in the same way. PCA provides strong evidence that metabolite pools are genetically correlated, although we cannot statistically assess the strength of correlations between individual metabolites. Some metabolites are also significantly correlated with fitness. Given this structure, we hypothesized that the sensitivity of a metabolite to mutation might depend on its position in the metabolic network. To test that hypothesis we reconstructed the C. elegans metabolic network from genomic data, but find that its topology does not predict either the rate of accumulation of mutational variance or the direction and/or magnitude of the mutational bias. We suggest that the observed patterns depend on higher level regulatory interactions that are not captured by simple metabolic network properties.