This study presents an analysis of the rate of dilution of a conservative nonreactive tracer in two well-known field experiments: The Borden (Ontario, Canada) experiment and the Cape Code (Massachusetts) experiment. In evaluating the dilution of injected sodium bromide, in addition to computing the second spatial moments, we have used the dilution index and the reactor ratio. The dilution index is a measure of the formation volume occupied by the solute plume, and the reactor ratio is a shape factor, which measures how stretched and deformed the plume is. Unlike the second moments, which may go up or down during an experiment, the dilution index should increase monotonically. The results for both plumes were quite similar. After an initial period the dilution index increased linearly with time, which is macroscopically equivalent to transport in two-dimensional uniform flow. The reactor ratio was relatively constant during the period of the experiments. Their values, about 0.72 for the Borden test and 0.63 for the Cape Cod test, indicate that the Cape Cod plume was more stretched and deformed than the Borden plume. The maximum concentration, which is an alternative to the dilution index for quantifying dilution, was found to be more erratic and more susceptible to sampling error. ¿ American Geophysical Union 1994