The paper concludes that changing real user costs could be an important contributory factor in the house price rise of the late 1970s. They are less able to explain the data for the 1980s, when real interest rates increased and real user costs rose substantially but real house prices declined relatively little. The finding that house prices do not behave as efficient asset prices may be important for understanding the past decade, during which homeowners may have only slowly recognized that the rapid house price appreciation of the 1970s would not be repeated. Investor expectations of rapid house price appreciation through much of the 1980s can resolve the puzzle of why house prices did not fall further. This raises the possibility that investors may extrapolate the recent decline in real house prices and conclude that the real carrying costs of houses are particularly high, hence reducing the demand for housing and the level of house prices still further.