In intelligence investigations, such as those into reports of chemical- or biological-weapons (CBW) use, evidence may be difficult to assemble and, once assembled, to weigh. We propose a methodology for such investigations and then apply it to a large body of recently declassified evidence to determine the extent to which an attribution can now be made in the Yellow Rain case. Our analysis strongly supports the hypothesis that CBW were used in Southeast Asia and Afghanistan in the late 1970s and early 1980s, although a definitive judgment cannot be made. The proposed methodology, while resource-intensive, allows evidence to be assembled and analyzed in a transparent manner so that assumptions and rationale for decisions can be challenged by external critics. We conclude with a discussion of future research directions, emphasizing the use of evolving information-extraction (IE) technologies, a sub-field of artificial intelligence (AI).