In the context of paratuberculosis control programs, it is important that MAP-infected herds are detected with the least possible costs, but with sufficient certainty. In the present work, various cost-effective approaches for herd diagnostics of paratuberculosis in cattle were evaluated. The suitability of the diagnostic approaches for detecting herds with a low prevalence of MAP shedders was also examined. Herd diagnostics were evaluated in study 1 using pooled individual fecal samples of a random sample. An investigation of three herds with 300 animals each was simulated. Each herd was represented by 300 individual fecal samples previously tested for MAP. In order to reflect a different within-herd prevalence (1.00%; 2.00% and 4.33%), n MAP-positive fecal samples with n=3; 6; 13 and 300?n MAP-negative fecal samples were assembled. Ten random samplings were simulated for each herd by randomly selecting 80 samples from the 300 individual fecal samples, which were divided into pools of size 5 and 10. Pools with at least one MAP-positive sample were analyzed using bacterial culture and direct PCR. It was shown that there is a positive correlation between the bacterial density in a pool and the pool sensitivity (probability of detecting a pool with at least one MAP-positive fecal sample as a MAP-positive pool). No significant difference in pool sensitivity occurred when comparing pools of size 5 to pools of size 10. When using direct PCR, the pool sensitivity resulted in 47.0% for pools of size 5 and 45.9% for pools of size 10. The pool sensitivity of pools analyzed by bacterial culture did not deviate significantly from the pool sensitivity of pools analyzed by direct PCR; this was the case for both pools of size 5 and 10. It was shown that the combination of direct PCR and bacterial culture can significantly increase the sensitivity. If both laboratory methods were combined when analyzing pools of size 10, the herd sensitivity for analyzing pooled individual fecal samples of a random sample was between 67.3% and 84.8%. In herds with a low prevalence of MAP shedders, a single examination of pooled individual animal fecal samples from a random sample is not sufficient to detect the herd with high certainty. In study 2, the suitability of repeated environmental samples and pooled milk samples for MAP diagnostics at herd level was evaluated, since even when using these inexpensive diagnostic approaches, a single examination is not sufficient to identify herds with low prevalence with high certainty. In 36 dairy herds housed in free stalls in Thuringia (federal state of Germany), twelve on-farm visits were carried out over a period of one year. At each on-farm visit, a liquid manure sample as well as a boot swab sample and three environmental samples were taken from the milking area, the main alleyway and the lactating cow floor. All of these samples were analyzed individually using bacterial culture and direct PCR. A pool of liquid manure and the other three environmental fecal samples was also analyzed using culture and PCR. In addition, individual milk samples, taken in the context of twelve dairy herd improvement tests, were pooled monthly into pools of size 25 and 50 and tested for MAP-specific antibodies using ELISA. A Bayesian model was used to estimate the cumulative sensitivity and specificity for repeated tests, taking into account the covariances between the different tests and between the repeated tests. In herds with a prevalence of at least 2% MAP shedders, at least seven sampling events were necessary to achieve a sensitivity of 95% using environmental samples. In order to achieve a sensitivity of 95% in herds having a within-herd prevalence of less than 2%, a combination of different sample types was necessary in addition to repeated samplings. The present study showed that the detection of MAP-infected herds with a low within-herd prevalence using inexpensive testing schemes is not sufficiently reliable when testing herds only once. However, environmental samples and pooled milk samples can be used to monitor MAP-free dairy herds if testing is repeated. A combination of sample types and/or laboratory methods can increase the sensitivity. In suckler herds or farms with straw bedding, analysis of pools of ten individual fecal samples using bacterial culture and/or direct PCR can be used. In order to confirm that MAP-free herds have a high probability of freedom from paratuberculosis, the herds should be tested repeatedly with MAP-negative results.