When comparing calculated heating consumption in residential buildings assuming standard usage with standardized measured data, then the two typically does not fit. In fact, measured consumption may be a fraction only of what was calculated. The reason is direct rebound behavior by the inhabitants. The paper shows the importance of direct rebound through measured results. First the temperatures, recorded in daytime and sleeping rooms in a sample of dwellings, are commented. Then follows a discussion of the indoor temperatures found when calculated energy consumptions for heating were forced to give the same numbers as measured. Next, two small scale analyses of energy data gained in low-income estates are commented, followed by test results on direct rebound in two dwellings, one non-insulated, the other well insulated. These data prove that the benefits of direct rebound are much larger in non-insulated than in well insulated homes. That fact is used to construct a rebound curve, starting from the normalized consumption data gained in 964 houses. The paper ends by showing the effect of energy price on direct rebound.