On the basis of an IITA impact study, this paper demonstrates the importance of establishing plausible linkages between development cooperation interventions and the relevant changes observed on the ground when assessing impact. It also proposes standards for establishing plausibility in this context. Experience has shown that impact evaluators often try to prove or quantify impact on a highly-aggregated level. However, this paper contends that in most cases, attribution gaps caused by the existence of too many other significant factors make it impossible to isolate the effects of a single development intervention. The authors maintain that, for most development interventions, to require impact evaluators to establish more than plausible impact relationships would force them to gloss over much information and over-interpret the data.