A model assuming that a person's evaluation of a given housing attribute is determined by a weighted sum of his/her evaluations of its perceived consequences for activities and life values, and that the evaluation of a particular housing alternative is obtained by summing this value-fulfillment across all attributes, was applied to data obtained by interviewing the adult members of 43 Swedish households searching for a new dwelling. The model was quite successful in predicting preferences for housing alternatives. When predicting evaluations of the households' present dwellings and choices among housing alternatives, however, the model was considerably less successful. A number of possible explanations were offered for the latter finding.