One-bond (1)J(NH) couplings have been measured in N-15-enriched human ubiquitin and range from 91.1 to 95.6 Hz. Measurements have been carried out using two different methods and at H-1 frequencies of 360, 500, and 600 MHz. The best method yields a precision of ca 0.02 Hz, and permits reliable measurement of the small changes (<0.3 Hz) in (1)J(NH) Splitting that occur when the magnetic field strength is increased from 8.5 to 14 T. The dependence of the (1)J(NH) splittings on the strength of the static magnetic field originates from two sources : a dynamic frequency shift caused by interference of the N-15 chemical shift anisotropy and the N-15-H-1 dipolar coupling relaxation mechanisms, and a dipolar contribution caused by a small degree of alignment resulting from the anisotropic magnetic susceptibility of the diamagnetic protein. Best fitting of the measured data yields an orientation-independent decrease of 0.11 Hz in the (1)J(NH) splittings at 609 MHz relative to 360 MHZ; in perfect agreement with theoretical predictions for the magnitude of the dynamic frequency shift. When fitting the measured J values to the theoretical model, containing only the dynamic frequency shift and dipolar coupling contributions, the reduced error in the statistical F-test is smaller than one, assuming a 0.02 Hz rms error in the experimental (1)J(NH) splittings. This confirms that the random error in the measured data J(NH) values does not exceed 0.02 Hz, and that effects other than the dipolar coupling and dynamic frequency shift are not detectable. Dependence of the change in (1)J(NH) On the orientation of the N-H bond vector within the molecular frame yields experimentally determined axial and rhombic magnetic shielding susceptibility anisotropies of -2.1 x 10(-28) and 0.7 x 10(-28) cm(3)/molecule, respectively. A small improvement of the fit is observed when the amide proton is positioned at a distance above or below the C-i-1’-N-i-C-i(alpha) plane which is about five times smaller than the out-of-plane distance predicted by ab initio calculations on a dipeptide analog in vacuum.