Free air carbon dioxide enrichment (FACE) of crop canopies with an array of vertical pipe (VVT) emitters avoids interference of the natural micrometeorological environment (canopy air flow, solar radiation) that is encountered in flow-through or closed-circulation growth chambers. However, both horizontal wind speed and eddy diffusivity increase with height above the crop canopy and decrease with depth into the crop canopy which would result in losses of carbon dioxide to the atmosphere above and retention of carbon dioxide within the crop canopy. Therefore, this modeling study was conducted to predict the effects of carbon dioxide injection rate vs. height, as well as canopy geometry factors, on the vertical and along-wind horizontal distribution of carbon dioxide. A two-dimensional numerical model was used with profiles of horizontal wind speed and eddy diffusivities as inputs. Data from a grid-pipe, ground-level, area-source FACE study (1969-1971) by D. N. Baker, L. A. Harper, and others in Mississippi was used to calculate the vertical distribution of eddy diffusivities. Three VVP release methods up to a height of 2.5 m were investigated: uniform injection with height, wind speed weighted injection with height, and eddy diffusivity times wind speed weighted injection with height. The first model runs were with a canopy height of 1.8 m representative of cotton grown at Yazoo City, Mississippi in 1987 and 1988. The target carbon dioxide concentration at the center of the 22-m ring was 500 ppm. The uniform injection case predicted carbon dioxide concentrations greater than 600 ppm near the base of the canopy and only about 450 ppm near the top. The eddy diffusivity times wind speed simulation showed that too much of the carbon dioxide was injected above the crop canopy and was lost to the atmosphere above. The wind speed weighted injection gave the best prediction with close to 500 ppm near the top of the crop at the center of the ring. All simulations showed a decrease in carbon dioxide concentrations across the array from the upwind side to the downwind side ranging from 20 to 40 ppm carbon dioxide. These overall predictions are generally in agreement with experimental measurements.