A modified procedure (Kalis et al 1998) for culture of Mycobacterium paratuberculosis (Mptb) from bovine faeces, based on the method of Jørgensen for faecal samples from individual cows was used for the culture of pooled faecal samples from five cows each. In eleven dairy herds faecal samples from 733 cows were cultured both individually and pooled. Pooling of samples was performed age depending (strategically). Individual cultures demonstrated the presence of Mptb in six from these eleven herds, and in 43 from these 733 cows. The pooled faecal cultures detected Mptb in seven out of eleven herds and in 28 out of 151 pools. Six culture positive animals were not detected in pooled faecal cultures but on the other hand two pools in which no positive animals were detected by individual culture, were culture positive. If compared with individual culture as golden standard, sensitivity from the pooled faecal culture method was 86 % and specificity was 96 %. The average number of colonies in positive cultures of pooled faecal samples was 22 and the total number of colonies in the five corresponding individual cultures was 31. The high sensitivity of this pooled faecal culture method was unexpected because experiments with pooled faecal culture based on another culture method reported a sensitivity of only 38%, compared with individual faecal cultures (Vialard et al 1993). This decrease of the total number of colonies after pooling the samples was lower than expected because mathematically this decrease was expected to be 80% instead of 29% in pools containing only one positive sample. The conclusions were: (1) the effect of pooling of faecal samples on sensitivity of the method cannot be estimated without taking into account the method of culture and the method of pooling. (2) culture of strategically pooled faecal samples with the modified Jørgensen method proved to be a good alternative for culture of individual faecal samples to detect the presence of Mptb in dairy herds and leads to a significant reduction of costs and (3) the difference between the sensitivity of the described culture method and the method reported earlier in pooled faecal samples cannot be explained and further research is needed before a general advice about pooling can be given.