We examine the effect of magnetic field strength and the use of the isotropic rigid rotor model (when rotational motion is anisotropic) on the accuracy of rotational correlation times (τR) of native proteins in solution determined from measurements of 13C spin-lattice relaxation times (T1) of α carbons. If τR is about 10-8 s, then small experimental errors in measured T1 values at 14 kG can result in large errors in τR. We compare τR values obtained (with the use of the isotropic model) from T1 measurements at 14.2 and at 63.4 kG. For horse myoglobin, the two magnetic field strengths yield essentially identical τR values. In the case of human hemoglobin, the difference between the two τR values is slightly outside the estimated experimental error. For the relatively nonspherical protein molecules hen egg-white lysozyme and bovine serum albumin, the τR values obtained at 14.2 kG are about twice as long as those obtained at 63.4 kG. The use of the axially symmetric ellipsoid model for the analysis of α-carbon relaxation data is examined. This model does not provide a significantly better agreement between the correlation times obtained at 14.2 and 63.4 kG (for lysozyme and albumin) than the isotropic model.