Abstract A blend of hexitols, comprising 80 parts by weight of mannitol and 20 parts of sorbitol, has been used in place of a frying fat in roasting nuts. Even though physicochemical tests demonstrated no measurable deterioration of the hexitol components on continuing repetitive use, biological studies were carried out to establish the safety of the end product. The latter was the novel dry roasted peanut product; the control consisted of conventionally air roasted peanuts. Both products were ground to peanut butter consistency and then stabilized with added hydrogenated peanut oil to facilitate the feeding programs. Male and female rats were maintained on the two peanut products for four generations. It was found that rats subsisting on diets containing the regular peanut butter increased in weight at a somewhat greater rate than did those on the diet containing the special peanut product. Efficiency of food utilization was not adversely affected and overall weight gains were satisfactory. All rats appeared normal and healthy throughout the experiment. There was no morbidity and no diarrhea. Breeding performance was slightly better among those rats on the special peanut product diet. No pathology was observed at autopsy and there were no differences in organ weights among the rats on the two diets. Total liver lipid and liver cholesterol levels were somewhat higher among the control peanut butter‐fed rats, as compared to those on the “special” peanut butter diet. On the basis of (a) published related studies involving biological evaluations of the control peanut butter diet vs. the same adequate diet but containing no peanut components and (b) the present findings, it is concluded that no unknown factors of a toxic nature are present in peanuts roasted in the molten hexitol blend.