The endangered Hawaiian Coot (Fulica alai) is one of few native waterbird species remaining in Hawaii, USA. On the island of Oahu, 56 adults were captured and sexed using a molecular method, and sexual dimorphism, seasonal variation in shield size (possibly sexually selected), and use of morphometrics to predict sex as an alternative to molecular sexing were investigated. The sex ratio was significantly male biased: 39 males, 17 females (P = 0.003). No morphometric character assessed independently differed significantly between sexes (0.056 < P < 0.975). However, a multiple regression approach indicated that tarsus length, bill height, tail length, and wing length in combination differed between sexes, but prediction of sex was nonetheless possible with only 65% accuracy. Male shield size was larger during the pre-breeding and breeding seasons than the post-breeding season, but did not vary significantly in females, the interaction between sex and season in the model being significant (P < 0.02). Within the breeding and pre-breeding seasons (a subsample of total birds), multiple regression indicated sexual dimorphism in shield width, shield length, and wing length, but prediction of sex was again only possible with 68% accuracy. Thus, neither model predicted sex sufficiently accurately for this morphometric approach to be a viable substitute for molecular sexing.