Mycobacterium avium subsp. paratuberculosis (Map) is a fairly prevalent disease in the southern part of Chile. For farmers that want to control the disease it is expensive to detect the infected cows with diagnostic tests. One way to decrease the costs is to use pooled fecal culture instead of individual tests.The objective of the study was to compare the validity and costs of pooling with individual fecal culture and an individual ELISA.Individual fecal and blood samples were taken from 50 cows >2 years old in 12 herds located in the south of Chile. In total, data on 598 dairy cows were obtained. Blood samples were tested for antibodies with an ELISA. The fecal samples were cultured for 16 weeks on solid media (HEYM). Both individual samples and pools of 5 and 10 animals were formed and cultured. Colonies suspected to be Map were confirmed with a PCR (IS900). The sensitivity of the pools to detect shedding cows was compared with the sensitivity of individual culture and ELISA results. Costs of pooling were compared with costs for individual fecal culture and for individual ELISAs.In total, 7% (42) individual samples were culture positive for Map. The pools with 5 and 10 animals detected 43.2% and 46.4% of the culture positive pools, respectively. The sensitivity differed for the level of shedding (higher for pools with high shedders than for pools with low shedders) and with the prevalence within a herd (higher sensitivity in herds with a higher prevalence). Pooling had a similar sensitivity as an individual ELISA; the moderate to high shedders were detected. However, pooling considerably decreased the costs, which should be weighed against a lower sensitivity compared with individual culture. The optimal testing strategy depends on the prevalence in a herd, the goal of the testing and the willingness to pay of the owner.