Recent studies indicate that houses with higher energy efficiency usually have higher market prices, a 'market premium for energy efficiency'. But in Germany the usefulness of this premium is confounded by the 'prebound effect': the gap between officially certificated energy ratings and actual energy consumption. Attempts have been made to close this gap from two complementary directions: downwards, by obtaining more accurate and less pessimistic technical estimates of idealised energy performance; and upwards, by estimating how much energy occupants realistically need for health and comfort. This study investigates prebound effects alongside an analysis of house prices in Germany, using a large database of house sale advertisements from 2007 to 2021, focusing on pre-1980 homes that were re-sold in 2019-2021. It uses ordinary least-squared multivariate regression to estimate market premiums for energy efficiency and sets these alongside estimates of prebound (and rebound) effects. It finds that prebound effects can lead purchasers to overestimate future energy savings and therefore pay more for properties than their actual worth. It also offers simple models to help purchasers interpret energy ratings more critically and estimate likely energy savings more realistically. Finally, it suggests how policymakers could modify energy ratings to reflect likely energy consumption more accurately.