Large multi-touch devices are gaining more attention in various elds of applications. The device through which the cartographic content is presented plays a role in cartographic interaction. This thesis presents a controlled user study comparing mouse- and touch-based map interaction on a large screen, in the context of target search tasks. Participants' mouse and touch input were logged in the browser. Several measurements were used to evaluate the difference between the mouse and touch interaction - drag length, drag direction, drag and zoom frequency, touch/mouse on-screen locations, and time spent on tasks. The results suggest that when using touch, users pan the map differently from when they are using mouse, with shorter panning distances and different main panning directions. The touch locations concentrate in the lower-middle part of the screen while mouse covers a larger screen area. Drag and zoom frequencies of the two input methods are quite similar, as well as the task completion times. In situ observation of user behavior identified several usage patterns and frequent problems when using touch. User feedback shows a clear preference of mouse input and gives reasons for participants' preference.