The mitochondrial oxidative theory of aging has been repeatedly investigated over the past 30 years by comparing the efflux of hydrogen peroxide (H2 O2 ) from isolated mitochondria of long- and short-lived species using horseradish peroxidase-based assays. However, a clear consensus regarding the relationship between H2 O2 production rates and longevity has not emerged. Concomitantly, novel insights into the mechanisms of reactive oxygen species (ROS) handling by mitochondria themselves should have raised concerns about the validity of this experimental approach. Here, we review pitfalls of the horseradish peroxidase/amplex red detection system for the measurement of mitochondrial ROS formation rates, with an emphasis on longevity studies. Importantly, antioxidant systems in the mitochondrial matrix are often capable of scavenging H2 O2 faster than mitochondria produce it. As a consequence, as much as 84% of the H2 O2 produced by mitochondria may be consumed before it diffuses into the reaction medium, where it can be detected by the horseradish peroxidase/amplex red system, this proportion is likely not consistent across species. Furthermore, previous studies often used substrates that elicit H2 O2 formation at a much higher rate than in physiological conditions and at sites of secondary importance in vivo. Recent evidence suggests that the activity of matrix antioxidants may correlate with longevity instead of the rate of H2 O2 formation. We conclude that past studies have been methodologically insufficient to address the putative relationship between longevity and mitochondrial ROS. Thus, novel methodological approaches are required that more accurately encompass mitochondrial ROS metabolism.