Adult Humboldt Penguins (Spheniscus humboldti) are sexually monomorphic in feather coloration, therefore sexing by direct observation is difficult. To determine if subtle morphometric differences exist, we compared gender as determined by DNA analysis with logistic regression leave-one-out analysis of bill length, bill depth and flipper length measurements collected from free-ranging Humboldt Penguins inhabiting a coastal island in Chile. The final chosen model used bill length and bill depth, with no interactions, and correctly predicted the sex of 87.7% of the 756 individuals. Of the 368 females, 13.5% were misclassified as males, while 11.1% of the 388 males were misclassified as females. Including flipper length improved the predictive value of the model slightly. The analysis was not useful in determining the sex of a captive population of Humboldt Penguins of known sex because of excessive beak growth seen in captive penguins.