Ten fresh-market tomato (Lycopersicon esculentum Mill.) genotypes were evaluated for stability of fruit firmness, citric acid, soluble solids, β-carotene and ascorbic acid concentrations, sugar : acid ratio, color, N content, and dry weight when grown in nine environments. Linear relationships between the genotype means for a given trait and the mean for the trait in each environment were used as an indicator of stability. A stable genotype for a given trait was considered to possess a regression coefficient (b1) ⩽ a coefficient of linear determination (r2) > 0.50, a genotype mean above the grand mean (mean of all genotypes), and a nonsignificant deviation from regression mean square (S2d). Using these criteria, stability in the nine environments was shown by the fruits of the various cultivars as follows: ‘Flora-Dade’, ‘FTE-12’, and D76I27 for firmness; ‘Castlehy 1035’ and ‘Sunny’ for citric acid; ‘Walter’ for soluble solids concentration; ‘FTE-12’ for ascorbic acid concentration; ‘Hayslip’, ‘Walter’, and ‘Burgis’ for sugar : acid ratio; ‘FTE-12’ and ‘Hayslip’ for β-carotene concentration: ‘Flora-Dade’ and 827115-IBK for color a/b; ‘Castlehy 1035’ and ‘Hayslip’ for dry weight; and ‘Walter’ for N content. Stable genotypes are less sensitive to environmental changes and are more adapted to favorable and unfavorable conditions than unstable genotypes. No genotype was found to be stable for every fruit quality trait in the nine environments. Stability of fruit quality characteristics should be considered in tomato breeding programs to develop genotypes adapted to diverse environmental and management conditions.