A large-scale survey of commercially pasteurised cows' milk, and pasteurisation trials using commercial-scale plant (2,000 l/h with turbulent flow) and naturally infected milk, were carried out to determine the efficacy of HTST pasteurisation applied to milk naturally infected with Mycobacterium avium subsp. paratuberculosis (MAP). During the milk survey a total of 567 commercially pasteurised milk samples from approved dairy processing establishments throughout the UK were tested for the presence of viable MAP. In the pasteurisation trials, naturally infected milk from two local farms was subjected to four different heat treatments (73ºC for 15 and 25 s, with and without prior homogenisation) on twelve separate occasions. In both studies, milk samples were subjected to decontamination with 0.75% (w/v) cetylpyridinium chloride for 5 h at room temperature before culture on Herrold's egg yolk medium and in BACTEC 12B medium. Overall, viable MAP was cultured from 10 (1.8%) of 567 pasteurised milk samples tested during the UK milk survey and from 10 (6.7%) of 144 pasteurised milk samples tested during the commercial-scale pasteurisation experiments. Surviving MAP were isolated from pasteurised milk samples that had been heated at 72ºC for both 15 and 25 s. The results of these two studies provide clear evidence that MAP bacteria in naturally infected milk are capable of surviving commercial HTST pasteurisation on occasion. It should be noted that the above studies differ from previous laboratory-scale pasteurisation studies in several respects: no MAP were artificially added to milk prior to heat treatment, a larger volume (50 ml) of pasteurised milk was tested thereby increasing the chances of detecting survivors, and 24-72 h elapsed between heat treatment and milk testing potentially giving sub-lethally heat-injured MAP cells time to recover viability.