We aim to contribute to the development of tactilebasedpedestrian navigation systems that help users to navigate urban environments with minimal attention to the userdevice interface. This paper describes the design and evaluation of a prototype and reports findings from (i) a labbased study that directly compared features of two widely researched forms of tactile display: a waist belt and a back array; and (ii) a field evaluation which compared our prototype tactilebased navigation system (TactNav) with a visual mobile maps application (Nokia Maps™). Lab results indicated that the waist belt afforded significantly better performance than the back array across a wide range of metrics. Field results indicated that users' performance with the tactilebasedsystem was equivalent to that with the visualbasedsystem in terms of accuracy while route completion time was significantly faster with the tactilebaseddirectional display.