The paper reports on the experiences of designing the interface of a mobile learning game. Three system versions were presented to discuss design issues and the results of three trials (the largest with more than 100 users). Eight design implications resulted from those trials: 1. Base the navigation interface on maps designed for walking. 2. Precise zooming is adequate for standing use only; button based coarse zooming is adequate for walking. 3. Good navigation support does not only cover the current situation, but also captures the past and supports planning. 4. Visualise imprecision of location information appropriately. 5. Chat-based communication is inadequate for many game situations. Carefully designed oral communication would be more appropriate in many situations. 6. Embed competitive awareness into the other channels. 7. Design explicitly for standing and walking use. 8. Design systems for user pairs. The design approaches discussed in this paper can also be applied to other mobile applications such as mobile learning in museums, emergency support or tourism.