Oxidant stress has been widely implicated as a mechanism of disease, yet clinical trials of antioxidants have not included a biochemical basis for dose selection or patient inclusion. Many of the indices traditionally employed to assess lipid peroxidation have relied on measurements performed in ex vivo systems of questionable relevance to events in vivo. Commonly employed in vivo indices of lipid peroxidation are constrained by such issues as the nonspecificity or instability of the target anylate, contamination of the anylate by events ex vivo, and nonspecificity of analytical methodology. More recently, specific methodology based on mass spectrometry has been applied to both 4-hydroxynonenal and a variety of isoprostanes in human biological fluids. Measurement of these compounds in urine reflects lipid peroxidation in vivo and offers a noninvasive approach that may be readily applied to clinical trials.