We report results on the performance of the combination of soft keyboards and marking menus. A model of expert user performance indicated an 11 - 37% (depending on the keyboard layout) improvement in text entry rate over the same keyboard without the menu. To verify the advantage in real usage, we conducted two experiments using the QWERTY keyboard layout with and without the menu. The first experiment imitated nearly perfect cognitive performance and measured motor performance. Using the menu saved time. The second experiment measured performance in a realistic text entry task. Initially using the menu slows down text entry. By the end of the 20-session experiment both conditions were equally fast. With continued practice text entry is likely to be faster with the menu. Categories and Subject Descriptors H5.2 [Information Interfaces and Presentation]: User Interfaces - Evaluation/Methodology, Input Devices and Strategies. Keywords text entry, Fitts' law, marking menu, soft keybo...