While many educational practitioners and researchers can speak to the benefits of teacher leadership anecdotally and theoretically, tracing direct evidence of impact via traditional quantified measures, like student achievement, has been challenging. In this paper, we argue that the logic underlying traditional teacher leadership assessment does not correspond with the complex realities of teacher development and, as such, cannot provide a meaningful assessment of teacher leadership programs. Drawing on complexity theory -- the study of complex, adaptive systems -- we suggest that a focus on information, and the processing of non-linear information sharing/exchange, may be a more fruitful analytic tool for evaluating the quality and effectiveness of teacher leadership efforts.