We examined sexual dimorphism in 11 morphological characters of the Flesh-footed Shearwater (Puffinus carneipes) and used these characters in a discriminant function analysis (DFA) to enable sexual classification in the field. We also used molecular techniques to determine gender in live birds. Sexual dimorphism was present in all characters tested, with males (n = 50) significantly larger than females (n = 52) in all cases. A stepwise DFA of the 11 morphological characters resulted in gender being correctly assigned in 91% of birds using head–bill length (HBL), minimum bill-depth (MBD), and superior unguicorn-width (SUW), using the generalised function –48.360 + (0.661 × SUW) + (0.931 × MBD) + (0.334 × HBL). Of the nine misclassifications, seven were a result of females being incorrectly assigned as males, indicating an overlap between the upper cut off value for females with the lower threshold value for males. We applied the DFA to an existing morphometric dataset (n = 219) of Flesh-footed Shearwaters caught as incidental by-catch from long-line fishing in the Eastern Tuna and Billfish Fishery. The generalised discriminant function derived from live birds performed poorly when applied to the by-catch sample (only 68.5% correctly sexed), and the issues of data consistency and the application of discriminant functions between populations are discussed. A two-character DFA constructed from the by-catch dataset (–30.128 + 0.206 × HBL + 0.848 × MBD) resulted in 86% of by-catch birds being correctly sexed. The findings suggest that a combination of the morphological traits HBL, MBD and SUW are sufficiently robust to use in the field to determine the gender of live birds. The merits and application of this technique are discussed in relation to seabird conservation and management.