|Plasma amino acid analysis is important for diagnosis and treatment of inborn errors of metabolism, and increasingly for the evaluation of nutritional status and metabolic derangement of other diseases. Its increasing clinical application is a result of the development of highly sensitive and reliable analytical methods. The interpretation of plasma amino acids is complicated and requires experience and expertise, not only because of the clinical and analytical complexities, but also because of the preanalytic variation on plasma amino acids. Many sources of preanalytic variations either remain unknown or are frequently ignored in practice, thus significantly compromising the usefulness of results. To better understand the preanalytic variations of plasma amino acids, this study started with review of the literature in current research on preanalytic variation. Topics included were blood sampling strategies, anticoagulant use, blood separation, deproteinization, storage, hemolysis, and buffy-coat contamination. This study, then, used both clinical plasma and blood specimens drawn from healthy adult donors to study the effects of delayed separation, hemolysis, and buffy coat contamination on plasma amino acids. In this study, the ratio of arginine to the sum of arginine and ornithine was discovered to be useful in identifying delayed separation of specimens. Using this ratio as a tool, the incidence of delayed separation in plasma homocysteine specimens was found to occur frequently. Effects of hemolysis were found to be variable and complicated by other factors, such as delays in separation, buffy coat contamination, and sample handling at room temperature. Quantitative determination of hemoglobin by spectrophotometer is simple and can be sued as a marker for assessing plasma sample quality. Preanalytic variation will continue to be an important issue for plasma amino acid analysis. The study of it effects will improve the quality of specimens and subsequently generate meaningful results.