Reed (1980) found that low islands and atolls in the tropical Pacific have a significant effect on measured rainfall. But direct examination of island data leads to the opposite conclusion. Reed and Elliott's (1979) rainfall estimates, the basis of Reed (1980), overestimate dry tropical islands and underestimate wetter islands by a factor of 2 or 3. Further, Reed and Elliott increasingly underestimate the tropical peak toward the west. Other observations are cited, such as satellite cloud analysis, that conflict with Reed (1980). It is concluded that the Reed (1980) analysis is flawed, and that the Reed and Elliott (1980) estimates are in error.