Sales records on boars sold from a breeding company and semen sold from a boar artificial insemination stud were evaluated to assess the emphasis placed on performance data and costs by buyers of semen and boars, and to determine what affects boar semen price. Two data sets, that consisted of (1) numbers of boars sold in a given breed and genetic evaluation subclass and (2) units or dollars of semen sold for a boar over a specific sales period, were used. These sales were expressed as a percentage (market share) of the total sales over the 2-yr period. Equations to predict percentage market share (% MS) for boar sales included the fixed effects of period and breed and covariates for boar price and genetic evaluation within breed. All effects were important (P<.01), and R² was .83. Percentage MS for units of semen sold was estimated using a model that included the fixed effects of period and breed and the covariates for cost per unit (CPU) of semen and backfat and gain within breed. Individual performance was more useful in predicting % MS and CPU than was any index. Subjective scores on mothering ability, size, soundness and libido were useful in predicting % MS units and CPU. For predicting % MS units, R² values were as high as .44 for models with scores. Results indicated that buyers were willing to pay more per unit and purchase more volume of semen from boars that had test information.