Study findings clearly show that little valuable information was produced from resemblance ratings generated from a five-point scale using a side-by-side comparison of a single face to a facial approximation, as is typically used in current facial approximation procedures. These findings add support to past claims that commonly used resemblance rating protocols offer little useful information about the accuracy of a facial approximation. Resemblance rating tests fail to produce results that are highly consistent with recognition performances; and they produce similar and moderately high resemblance scores irrespective of which face is used for the comparison. This is particularly evident when comparison faces are presented singly and in isolation. The findings of this study and similar studies reported in the literature do not rule out the possibility that resemblance rating scales may be useful with some alteration, but the findings strongly suggest that currently used methods are, at least, insensitive. This paper describes the materials and methods used in the two experiments, one in which face array and resemblance rating tests used the same facial approximation and another which used resemblance rating scores of disparately recognized foil faces. 6 figures and 48 references