Systematic errors in time reproduction tasks have been interpreted as a misperception of time and therefore seem to contradict basic assumptions of pacemaker-accumulator models. Here we propose an alternative explanation of this phenomenon based on methodological constraints regarding the direction of time, which cannot be manipulated in experimental settings. In two experiments, we demonstrate the influence of the direction of a dimensional change for pitch and brightness estimates. The results support the assumption that errors in time reproduction tasks do not reflect a systematic temporal misperception, but rather a methodological artifact. Implications for contemporary models of time perception are discussed. (Contains 3 figures and 4 footnotes.)