Each seed phenotypic trait may be influenced by its metabolite composition which varies during seed development and -maturation. Therefore, the metabolic components of dry seeds may represent the maternal environment under which seeds developed and matured. The natural variation of seed metabolite composition has been previously studied and several QTLs were detected regulating this variation. Here a generalized genetical genomics (GGG) approach was used for the metabolic analysis of a recombinant inbred (RIL) population obtained from a cross between two tomato species: Solanum lycopersicum and Solanum pimpinellifolium. The RILs and parental lines were grown in two maternal environments: high phosphate and low nitrate nutrition. A correlation analysis of metabolite composition and seed phenotypic traits indicated several relations between metabolite contents and seed quality traits such as seed size, seed weight and seed germination percentage. Seed size and -weight exhibited a positive correlation with several amino acids and some intermediates of the TCA cycle, such as succinate, citrate and malate. By performing metabolic correlation analysis and also generating metabolite networks and combining these with QTL analysis, several relevant mQTLs were identified. We showed that dry seed metabolites and QTLs regulating them are modulated by genetic factors, maternal environment and their interaction.