A series of tracer tests has been conducted in a 7-m-thick fractured dolomite at two sites in southeastern New Mexico. The tests were designed to evaluate transport processes, especially matrix diffusion, in fractured, permeable media. Both single-well injection-withdrawal (SWIW) and multiwell convergent flow (MWCF) tests were conducted. Seventeen different organic tracers (the fluorobenzoic and chlorobenzoic acids) and iodide were used as conservative tracers for the tests. The MWCF tests included repeated tracer injections while pumping the central well at different rates, injection of tracers with different aqueous diffusion coefficients, and injection of tracers into both the full and partial formation thickness. This paper describes the tracer test sites and aquifer characteristics, the experimental methods, and the tracer data produced. The tracer test results provide a high-quality data set for a critical evaluation of the conceptual model for transport. Both the SWIW and MWCF tracer test data showed gradual mass recovery and breakthrough (or recovery) curve tailing consistent with matrix diffusion. However, the SWIW recovery curves did not display the -1.5 log-log slope expected from a conventional double-porosity medium with a single rate of diffusion. The breakthrough curves from MWCF tests conducted at two different pumping rates showed similar peak heights, which is also not what was expected with a conventional double-porosity model. However, the peak heights were different for two tracers with different aqueous diffusion coefficients that were injected simultaneously in one test, consistent with the effects of matrix diffusion. The complexity of the tracer test results suggests that a simple double-porosity conceptual model for transport in the Culebra with a single rate of diffusion is overly simplistic. ¿ 2001 American Geophysical Union