15N NMR studies of the coliphage lambda cro repressor are presented. The protein has been uniformally labeled with 15N, and individual amino acids have been incorporated. Although the four C-terminal residues (63-66) were not located in the original crystallographic studies of the protein [Anderson, W.F., Ohlendorf, D.H., Takeda, Y., & Matthews, B.W. (1981) Nature (London) 290, 754], it has been proposed that the C-terminus is involved in DNA binding [Ohlendorf, D.H., Anderson, W.F., Fisher, R.G., Takeda, Y., & Matthews, B.W. (1982) Nature (London) 298, 718]. These experiments give direct verification of that proposal. [15N]Amide resonances are assigned for residues 56, 62, 63, and 66 in the C-terminus by enzymatic digestion and by 13C-15N double-labeling experiments. 15N[1H] nuclear Overhauser effects show that the C-terminus is mobile on a nanosecond time scale. Exchange experiments using distortionless enhancement via polarization transfer, which is sensitive to proton exchange on the 1/JNH (10 ms) time scale, indicate that the amide protons in the C-terminus are freely accessible to solvent. It is thus a flexible arm in solution. The binding of both specific operator and nonspecific DNA is shown to reduce both the mobility and the degree of solvent exposure of this arm. Two-dimensional 15N-1H correlation experiments using 15N-labeled cro reveal inconsistencies with previously reported 1H NMR assignments for the lysine amides [Weber, P.L., Wemmer, D.E., & Reid, B.R. (1985) Biochemistry 24, 4553]. This result suggests that those assignments require reexamination, illustrating the utility of 15N labeling for obtaining 1H resonance assignments of biomolecules. Furthermore, isomerization of the peptide bond of Pro-59, which has been previously suggested (Weber et al., 1985) and which would significantly affect the properties of the C-terminal arm, is shown to not occur.