Two forced‐gradient ground‐water transport tests were conducted in the Summers of 1989 and 1990 at Borden, Ontario. The tests were nearly identical in several respects and allow an assessment of the reproducibility of such efforts. In both experiments, carbon tetrachloride, tetrachloroethene, and a conservative tracer were injected as a pulse into the aquifer and monitored as they migrated towards an extraction well. Data were analyzed using moment techniques and fitted with a 1‐D transport model accounting for physical nonequilibrium conditions for the sorbing solutes. Visual inspection and simulations of the results indicate that the tests are reproducible. However, results from the moment analyses are less conclusive, perhaps due to data truncation. The differences in the modeling results or the parameter estimates derived from them are likely to have resulted from differences in the experimental conditions or duration. Simulations conducted using either year's fitted parameters are essentially indistinguishable from a practical point of view. Copyright © 1994, Wiley Blackwell. All rights reserved