Those who argue that dispositional explanations are genuine explanations usually construe them as causal explanations. There are several well-known arguments against the causal efficacy of dispositions, but there are as well demonstrations that on some minimal conditions, dispositions could be viewed as causally relevant to the effects which they are taken to explain. Although the latter position is generally tenable, it may be shown that in some important cases it is not a good idea to commit to a causal construal of dispositional explanations. The argument goes as follows: (1) Dispositional explanations are valued for certain specific extra-inferences which they allow us to draw; (2) The causal construal of dispositional explanations can account for some of these extra-inferences only on the assumption that the disposition is a common cause of its manifestations; (3) However, under certain circumstances, the common cause assumption is refuted on theoretical or empirical grounds; Therefore, (4) under certain circumstances, the causal construal of dispositional explanations cannot account for what these explanations are valued for. The latter conclusion is a reason to argue that in some cases at least, the causal construal of dispositional explanations should be avoided.