A probabilistic model was validated for estimating avian mortality associated with the application of the avicide CPTH (3-chloro-4-methylaniline hydrochloride) to minimize sprouting rice damage in the southern United States. CPTH exposures for individual birds were predicted by random sampling from species-specific non-parametric distributions of bait seed consumption and CPTH residues detected on individual bait seeds. Mortality was predicted from the species-specific exposure versus mortality relationship. Individual variations in this response were captured in the model by Monte Carlo sampling from species-specific distributions of slopes and median toxicity values (LD50) for each bird. The model was used to evaluate the effects of formulation, bait preparation and application procedures on target and non-target mortalities. The results of these analyses indicate that: 1) decreasing the concentration of CPTH on the treated bait seeds from 400 to 300 μg CPTH/seed will improve bait performance by decreasing non-target mortalities, 2) the current dilution ratio of 1 treated seed to 25 diluent seeds is optimal and 3) preparing a bait seed product in which CPTH is homogeneously distributed throughout the bait seed mixture will significantly increase bait performance by increasing red-winged blackbird (target) mortality and decreasing mortality for savannah sparrows and meadowlarks (non-targets).