The rate of consumption of dithiothreitol (DTT) is increasingly used to measure the oxidative potential of particulate matter (PM), which has been linked to the adverse health effects of PM. While several quinones are known to be very reactive in the DTT assay, it is unclear what other chemical species might contribute to the loss of DTT in PM extracts. To address this question, we quantify the rate of DTT loss from individual redox-active species that are common in ambient particulate matter. While most past research has indicated that the DTT assay is not sensitive to metals, our results show that seven out of the ten transition metals tested do oxidize DTT, as do three out of the five quinones tested. While metals are less efficient at oxidizing DTT compared to the most reactive quinones, concentrations of soluble transition metals in fine particulate matter are generally much higher than those of quinones. The net result is that metals appear to dominate the DTT response for typical ambient PM_2.5 samples. Based on particulate concentrations of quinones and soluble metals from the literature, and our measured DTT responses for these species, we estimate that for typical fine particle samples approximately 80% of DTT loss is from transition metals (especially copper and manganese), while quinones account for approximately 20%. We find a similar result for DTT loss measured in a small set of PM_2.5 samples from the San Joaquin Valley of California. Because of the important contribution from metals, we also tested how the DTT assay is affected by EDTA, a chelator that is sometimes used in the assay. EDTA significantly suppresses the response from both metals and quinones; we therefore recommend that EDTA should not be included in the DTT assay.