Abstract In theory, residential mobility is a response to environmental stress only if households do not reduce dissatisfaction through other alternatives, such as housing improvements or repairs. Despite the attention given to stress-reducing alternatives, however, no attempt has been made to test empirically the residential satisfaction model with adjustments. Using data from the Annual Housing Survey: 1978–1981, I model three stages in the mobility process and investigate potential sources of specification error in previous tests. Blocks of family cycle, background/action state, and location/housing variables are shown to affect adjusting significantly. Residential satisfaction strongly affectsmobility preferences; and all theoretically relevant blocks of explanatory variables predict mobility. Alternatives to mobility should be included in the residential satisfaction model.