A GC-based approach was applied to compare the metabolite profiles of two low phytic acid (lpa) soybean mutants and their respective wild-types. The lpa mutants (Gm-lpa-TW75-1 and Gm-lpa-ZC-2) were grown together with the wild-types (Taiwan 75 and Zhechun no. 3) in three and four field trials, respectively. HPLC analysis revealed a phytic acid reduction of -53% for Gm-lpa-TW75-1 and of -46% for Gm-lpa-ZC-2. For Gm-lpa-TW75-1, no accumulation of lower inositol phosphates was observed, whereas Gm-lpa-ZC-2 exhibited significantly increased contents of the lower inositol phosphates InsP(3), InsP(4), and InsP(5) compared to the corresponding wild-type. The metabolite profiling revealed that compared to the wild-types, 40% (Gm-lpa-TW75-1) and 21% (Gm-lpa-ZC-2) of the detected peaks were statistically significantly different in the lpa mutants grown at one field trial. However, the majority of these differences were shown to be related to environmental impact and natural variability rather than to the mutation event. Identification of consistent metabolic changes in the lpa mutants revealed decreased contents of myo-inositol, galactinol, raffinose, stachyose, and the galactosyl cyclitols galactopinitol A, galactopinitol B, and fagopyritol B1 compared to the wild-type. These consistently pronounced changes in Gm-lpa-TW75-1 confirmed the suggested mutation target. Consideration of the metabolic changes observed for Gm-lpa-ZC-2 (accumulation of lower inositol phosphates and increased myo-inositol contents) indicated a mutation event affecting the latter biosynthetic steps leading to phytic acid. The study demonstrated the applicability of metabolite profiling for the detection of changes in the metabolite phenotype induced by mutation breeding and its power in assisting in the elucidation of mutation events.