The objective of this paper is to determine how microreactor geometry affects performance. In particular we consider how the characteristics of a posted microreactor and a channel microreactor are different for a test reaction: the production of hydrogen from ammonia. The experiments include kinetics measurements, using a recycle reactor to obtain an empirical expression for the rate of reaction under the experimental conditions, and rate measurements to verify performance. We find that the performance characteristics of the two reactors are quite different. In the range of conditions studied, the conversion in the posted reactor is quite similar to that for an ideal CSTR, while the performance of the channel reactor is quite similar to that for an ideal PFR.