Are behavioral economics games reliable tools for measuring ethnic biases? Empirical experiences with these tools, together with their vulnerability to self-presentational concerns and insensitivity to more automatic associative or emotional reactions, raise questions as to what biases they might not detect. We recruit experimental subjects in Nairobi, Kenya, testing their behavior in a set of common behavioral games together with their performance on a set of social psychology tasks designed to both resist social desirability bias and to capture a range of automatic associative and emotional reactions participants experience in response to non-coethnics. Our results suggest that there are ethnic biases within this group and context, likely reflecting positive/negative feelings and threat perception toward non-coethnics, that standard behavioral games fail to detect. We conclude that researchers interested in measuring ethnic biases of various types should consider adding the social psychology tasks to their tool kits.