Convergent flow tracer tests conducted in the Culebra dolomite (Rustler Formation, New Mexico) are analyzed with both single- and multiple-rate, double-porosity models. Parameter estimation is used to determine the mean and standard deviation of a lognormal distribution of diffusion rate coefficients as well as the advective porosity and longitudinal dispersivity. At two different test sites both multirate and single-rate models are capable of accurately modeling the observed data. The single-well injection-withdrawal test provides more precise estimates of the mass transfer parameters than the convergent flow tracer tests. Estimation of the multirate distribution parameters is consistent across locations for the two types of tests. Limits of resolution are calculated for the multirate distribution, and these limits explain the precision with which the standard deviation of the multirate distribution can be estimated. These limits also explain the necessary increase in the advective porosity for the single-rate model at one location and not the other. Implications of the multirate mass transfer model at time and length scales greater than those of the tracer tests include the instantaneous equilibrium of a significant fraction of the matrix and the possibility of a fraction of the diffusive porosity not reaching an equilibrium solute concentration at long times. ¿ 2001 American Geophysical Union