An experiment is reported that contrasted two recent hypotheses about the guiding effects of knowledge of results (KR) on motor learning. Specifically, the purpose was to compare the impact of the nature of the KR itself versus the effect of the information conveyed by the KR on learning. We compared four groups of subjects, comprising a factorial combination of KR conditions (specific or bandwidth) and the movement timing goal (specific or bandwidth). All subjects performed 100 acquisition trials and 20 retention trials. The results revealed that bandwidth KR conditions facilitated learning, regardless of the specificity of the movement goal. These effects support the hypothesis that when error KR is provided, its very provision can block error detection activities that are important for retention performance.