Bromide breakthrough curves from push-pull tests were obtained at two wells before, during, and after citrate injections to assess how citrate-induced colloid mobilization affected physical aquifer transport properties. Tailing and incomplete bromide recoveries (67--95%) could not be fit with a conservative advection/dispersion model, and the results of batch tests using aquifer solids implied bromide was not significantly sorbing. Thus we modeled the bromide returns considering advection, dispersion, and rate-limited diffusive mass transfer between mobile and immobile regions by fitting αr, the radial dispersivity; α, the rate-limited mass transfer coefficient; and ¿, the volumetric ratio of immobile-to-mobile domains. Statistical t-tests indicated that the changes in aquifer transport parameters at a well where colloid mobilization was limited were not significant at a 95% percent confidence level. However, the substantial colloid mobilization at a second well corresponded to significantly decreased αr and ¿, while increasing α between premobilization and both mobilization and postmobilization. The changes in aquifer parameters and their correlation to the recovered colloidal mass are consistent with the idea that pore-clogging colloids were mobilized and/or reorganized during citrate injections. The results suggest that flushing a site under the right conditions with citrate could open up immobile regions and substantially reduce remediation time and costs by liberating contaminants whose transport would otherwise be diffusion limited.