Abstract A suite of 13C-detected NMR pulse sequences to edit the correlation peaks of the CACO and CON spectra according to the amino acid residue type is presented. The pulse sequences exploit the topology of the Cβ carbon and led to the sorting of the CACO or CON signals into several classes depending on the nature of the generating residue. A set of four or eight correlation spectra is recorded where the sign of the cross peaks changes from one spectrum to another according to the amino acid type of the corresponding residue in the protein sequence. Linear combination of these spectra produces subspectra showing signals from residues having similar Cβ topology. The presence of weak breakthrough peaks does not prevent the proper classification, since this is obtained from the subspectrum in which the correlation peak is more intense. The experiments were tested on a globular protonated protein (13C, 15N labeled Ubiquitin), on a globular deuterated protein (2H, 13C, 15N labeled Ubiquitin), and on an intrinsically disordered protein (13C, 15N labeled Nupr1).