Mining walking pattern from mobile users represents an interesting research area in the field of data mining which is about extracting patterns and knowledge out from a given dataset. There are a number of related works in knowledge extraction from mobile users, but none have previously examined the situation of how mobile users walks from one location of interest to another location of interest in the mobile environment. Walking pattern is the proposed method where it examines from the source data in order to find out the two-step, three-step and four-step walking patterns that are performed by mobile users significantly and strongly through location movement database using measure of support and confidence. Performance evaluation shows the tendency for the increased number of candidate walking patterns with the increase in location of interest and steps. Walking pattern has proven itself to be a suitable method in finding knowledge from mobile users.