The side chain of the serine residue in the active center of atropinesterase (AtrE), alpha-chymotrypsin (Chymo), and subtilisin A (Sub) was labeled with two paramagnetic reporter groups of different size (label I or II, respectively) by sulfonylation with N-[3-(fluorosulfonyl)phenyl]-1-oxy-2,2,5,5-tetramethyl-pyrroline-3 -carboxamide or N-[6-(fluorosulfonyl)-2-naphthyl]-1-oxy-2,2,5,5-tetramethylpyrroline+ ++-3 -carboxamide. ESR spectra of labeled enzymes in 10 mM phosphate buffer, pH 7.4, were measured at temperatures between 133 and 298 K by using a home-built spectrometer operating in the absorption mode at 10-kHz field modulation. The spectra, in particular those at 276-298 K, were analyzed by computer simulation of the overall line shape according to the methods developed by Freed and co-workers, based on eigenfunction expansion. In the case of AtrE for both labels, the best agreement between experimental and simulated solution spectra was obtained with only one mobility component showing anisotropic, axially symmetric reorientation according to the Egelstaff jump-diffusion model. The axis of preferential reorientation was found to lie in the XZ plane at a polar angle of about 30 degrees with the X axis. The corresponding rotational correlation time (tau parallel) did not show appreciable viscosity/temperature (eta/T) dependence but had a constant value of 4.4 and 2.2 ns for labels I and II, respectively. The rotational correlation time associated with rotation around the axes perpendicular to that of preferential reorientation (tau perpendicular) showed the usual eta/T dependence and had a value of 22.0 ns at 276 K for both labels. The above results strongly suggest that in AtrE both nonpolar reporter groups reside in a pocket near the active serine. Contrary to the situation in AtrE, the overall mobility of the -N-O. fragments in Chymo and Sub was found to result from contributions of at least two distinct motional states, strongly and weakly immobilized. In going from label I to label II, the relative contribution of the latter state increases at the expense of that of the former. This is ascribed to an equilibrium between a relatively free state of the aromatic cores and a firmly bound position in the specificity pocket of these proteases. The apparently more rigid embedding of the spin-labels in the enzyme structure of AtrE suggests that the size of the nonpolar binding pocket in the active center region of this esterase allows a deeper penetration of the aromatic portions of the labels than is possible for the specificity pocket of Chymo or Sub.