Nuclear magnetic relaxation data of water nuclei at variable fields provide valuable information on the dynamics of watersolute interactions. [1][2][3] However, information can be collected only within certain field ranges in which the nuclear relaxation rates are field-dependent owing to the dispersion of the spectral density, J(w,t). The dispersion depends on the type of motion and on the observed nucleus. The most informative 1 H NMR spectroscopic frequency range for rotational motions is centered around 10 MHz for small proteins (M W % 10 4 Da) and smaller frequencies for larger proteins and protein aggregates. Of course, resolution is low at these fields and in practice only one signal (or an unresolved signal envelope) can be detected. Furthermore, only the abundant water protons can be conveniently studied under such low sensitivity.Relaxation of the isotopes 17 O and 2 H in enriched water can also be studied; [1] in these cases the centers of the informative field ranges increase by a factor % 7 with respect to 1 H as a result of the magnetogyric ratios of 17 O and 2 H. Because water interacts with proteins, such relaxation studies provide direct information on the nature of water-protein dynamics, but give only indirect information on the protein itself.[4-6] It would be highly desirable to have complementary information directly from the protons of the protein, but this has been impractical so far owing to the low sensitivity of the available instrumentation. Protein 1 H relaxation data have been reported only on protein solutions of concentrations ! 35 % by mass, [7,8] far from physiological conditions. Improvements in field-cycling technology [9][10][11][12] have led to the production of instrumentation [13] with a % 10-fold increase in the signal-to-noise ratio, [14] with which direct protein 1 H detection can be attempted for protein concen- Figure 1. Protein 1 H relaxation dispersions for lysozyme solutions (2.8 mm in D 2 O) at a) pH* 3.5, and b) pH* 9.0 (pH=pH-meter reading in D20 solution). The time decay of the collective protein 1 H magnetization at 0.1 MHz and its single-exponential fit are shown in the inset of part a). Theoretical relaxation rates were obtained through singleexponential fits of time decays of collective protein 1 H magnetization calculated from the known protein structure of lysozyme [17] and a complete relaxation matrix (CORMA) analysis.[18] Only exchangeable NH protons from secondary structure elements ( % 50 % of total) were included, whereas all other exchangeable proton positions were assumed to be deuterated. Inclusion of either all or none of the NH protons affected the calculated rates by AE 2 %. The theoretical lowfield hE 2 i values obtained in this way were used to fit Equation (1) to the data (solid lines). The dotted lines represent data fits with two J-(w,t) terms in Equation (1) ( Table 1). Panel c) shows the protein 1 H relaxation dispersion for a solution of a-synuclein in D 2 O (1.4 mm, pH* 7.1). The solid and dotted lines represent fits with one or two J-(...