The effectiveness of Zdanovskii's rule as a fundamental method of predicting water activities in mixed aqueous strong electrolyte solutions has been investigated. Unfortunately, when literature data available for more than 100 ternary systems (i.e., two electrolytes plus water) were examined, sufficient measurements were rarely available to test Zdanovskii's rule conclusively. This is because few mixed systems have been studied by more than one (independent) investigator and, in these few cases, the differences between reported water activities are generally much larger than typical estimates of experimental uncertainty. Most of the electrolyte systems considered in this work are found statistically to conform with Zdanovskii's rule but often only within a worryingly large tolerance in residuals. However, when data are taken from a single source, deviations from Zdanovskii's rule, significant in comparison to the stated experimental precision, can often be attributed equally well to small, unrecognized, systematic errors. This dilemma can only be resolved by more experimental measurements of sufficient accuracy to achieve better agreement between independent observers. Our findings imply that (a) confidence in deviations from Zdanovskii's rule typically depends on a subjective judgment of data quality and (b) Zdanovskii's rule, albeit often difficult to validate, provides an impressively general method for estimating water activities in mixed aqueous strong electrolyte solutions, with a certainty that, objectively, is as good as or better than current experimental capability.