A recent experiment by Banerjee and Korpel [Phys. Fluids 25, 1938 (1982)] indicates that McGoldrick's [J. Fluid Mech. 21, 305 (1965)] theory of resonant interactions among deep water capillary-gravity waves may require modification. The conclusions are reexamined, and it is found that the experimental results are in overall agreement with the theory. Some systematic differences remain at low and high frequencies. Possible reasons for these discrepancies are proposed.