Personnel tour scheduling problems are highly relevant to service organizations and have recently received considerable attention in both theory and practice. The inherent complexity of generalized set-covering formulations (GSCFs) of personnel tour scheduling problems has often resulted in the deployment of LP-based and local-search heuristic procedures, even for relatively small problems. This paper evaluates the performance of the dual all-integer cutting plane for solving such problems. A computational study revealed that the cutting plane, enhanced by an LP objective cut and sophisticated source row selection rule, substantially outperformed a commercial branch and bound code across four sets of test problems. The study also showed that an advanced starting solution based on the LP relaxation of the GSCF generally provided more rapid convergence of the algorithm.