Previous research suggested that the chemical composition of rock varnish, a micrometer-thick coating of silicon, aluminum, manganese, and iron, changed predictably through time and so could be used as a chronometer. In particular, the cation ratio, (calcium + potassium)/titanium, was posited to decrease as the varnish aged. However, many earlier varnish studies were not well documented and were based on chemical analyses now shown to be inaccurate. This contribution describes a test and the first direct comparison between two different analytic methods used previously to cation-ratio date archaeological and geological materials. It provides methodology sufficiently detailed for replication of analytical and sampling techniques. The authors' analyses of rock varnish from a prehistoric chert quarry in California show that the two principal methods used to cation-ratio date rock varnish generate consistently different elemental abundances and are therefore not comparable. Neither analytic method produced a significant time-dependent trend in varnish chemistry, suggesting rock varnish cation ratios are not a reproducible and thus probably not a reliable way to date lithic artifacts.