The extent and location of transient structure in intrinsically disordered proteins (IDPs) provide valuable insights into their conformational ensembles and can lead to a better understanding of coupled binding and folding. Millisecond amide hydrogen exchange (HX) can provide such information, but it is difficult to quantify the degree of transient structuring. One reason is that transiently disordered proteins undergo HX at rates only slightly slower than the rate of amide HX by an unstructured random coil, the chemical HX rate. In this work, we evaluate several different methods of obtaining an accurate model for the chemical HX rate suitable for millisecond hydrogen exchange mass spectrometry (HX-MS) analysis of disordered proteins: (1) calculations using the method of Englander [Bai, Y., et al. (1993) Proteins 17, 75-86], (2) measurement of HX in the presence of 6 M urea or 3 M guanidinium chloride, and (3) measurement of HX by peptide fragments derived directly from the proteins of interest. First, using unstructured model peptides and disordered domains of the activator for thyroid and retinoid receptors and the CREB binding protein as the model IDPs, we show that the Englander method has slight inaccuracies that lead to underestimation of the chemical exchange rate. Second, HX-MS measurements of model peptides show that HX rates are changed dramatically by high concentrations of the denaturant. Third, we find that measurements of HX by reference peptides from the proteins of interest provide the most accurate approach for quantifying the extent of transient structure in disordered proteins by millisecond HX-MS.