The objective of the study was to investigate financial aspects of targeted sampling of subpopulations of cows on classification of herd paratuberculosis status, as part of level 1 of the Voluntary Johne's Disease Herd Status Program. All cows from 3 known infected herds with low to moderate prevalence, and from a non-infected herd in California were tested. The infection status of each cow was determined using ELISA results alone, or ELISA followed by fecal culture. Simulation methods were designed to randomly sample cows based on their lactation number, stage of lactation, and milk production. Sample sizes were ranged from 30 to 300 in herds that had 329 to 2474 cows. Simulation results showed that testing 30 cows in lactation ≥ 2 with ELISA followed by fecal culture detected 3% and 34% of paratuberculosis infected herds with estimated within-herd prevalences of 4% and 15%, respectively. ELISA testing alone provided a higher detection probability; however, it would misclassify 5%-89% of non-infected herds, depending on the number of samples collected. Sample size and costs could be decreased by up to 50% with targeted sampling of cows while achieving the same detection probability. In most situations, samples from cows in lactation ≥ 2 and DIM ≥ 200 days provided the highest detection probability. Findings from the study indicate that detection of Map infected herds could be improved, and cost of testing could be substantially reduced by sampling targeted groups of cows.