Five free-living women (ages 23-38 years) who consumed a self-selected diet and five women (ages 23-44 years) residing in a metabolic unit who were fed constant diet were assessed for variation in vitamin and general chemistry indices. Blood was drawn from these women once a month for five months, once a week for five weeks, and once a day for five days to assess analytical and biological variability of the indices. Analytical variability was determined by concurrently analyzing control samples prepared from plasma and serum pools. All samples were analyzed in duplicate. Of the measured indices, vitamins and lipids seemed to be the most variable. Diet had a significant effect only on ascorbic acid. We were unable to show any seasonal change for these analytes. Estimations of analytical variability, along with estimates of biological variability, and knowledge of dietary practices are essential when interpreting differences in analytes.