This paper analyses the determinants of, and mechanisms behind, local moving activity by existing households. This is done using a three-step procedure that is not used in prior studies of residential mobility. The procedure combines information on observed moving choices, stated causes of moves and moving plans. It can be said to be a combination of revealed determinants of mobility and stated causes and preferences. Among the results it can be mentioned that for non-movers an empirical logit-model of moving activity reveals a clear pattern where predicted moving probabilities predict moving plans fairly well. This is taken as support of a hypothesis that the decision to move is a process stretched out in time. Comparison of predicted moving probabilities and stated causes furthermore reveals that some types of moves are better captured by a general empirical model than others are. Hence there is a great diversity of possible mechanisms that can generate moves at an individual household level.Residential mobility, housing mobility, Norway, local housing markets, moving plans, stated and revealed preferences,