I recently stumbled across a great video tutorial for the OpenSesame experiment builder. This video, created by Chris Longmore, shows how to build an experiment in which participants have to judge the gender of a (picture of a) cat. This is kind of goofy, but experiments of this kind have been conducted at least twice. Once by Quinn and colleagues, and once by Longmore himself. And, apparently, people are able to distinguish male cats from female cats. Barely, but still. (You can participate in an online version of the experiment here.)
Here's the video:
And, because I'm sure you're wondering, here are some pictures of cats split by their actual and perceived gender. It seems (to me) that people go largely by colour and size: Big, dark cats look male, whereas slender and lightly coloured cats are perceived as females.
Quinn, P.C., Palmer, V., & Slater, A.M. (1999). Identification of gender in domestic-cat faces with and without training: Perceptual learning of a natural categorization task. Perception, 28(6), 749-763. doi:10.1068/p2884