We compare the recent pseudopotential calculation of potential curves for K-2 and NaK molecules [Phys. Rev. A 54, 204 (1996)] with the newest experimental data on selected molecular stares. It is shown that our experimental results reduce the discrepancies between theory and experiment.