This study demonstrates that a three-week stabilization period, in which all subjects eat an identical diet, produced a more uniform but different baseline of metabolic parameters than the subject's self-selected or“habitual”diets. Subjects required more food energy to maintain initial body weights during the stabilization period than when they ate their reported self-selected diets; average intakes of almost all nutrients were higher from the stabilization than from the self-selected diet. The switch to the stabilization diet produced small but significant reductions in blood pressure, in some serum enzymes, urine volume, and sodium; and statistically significant increases in serum LDL cholesterol, potassium, aldosterone, protein, albumin, phosphorus, BUN, and in urine potassium. The findings indicate that results must be interpreted with caution from studies in which the baseline for measuring metabolic variables is established by feeding subjects a standardized diet that differed markedly from their regular, self-selected diets.