Reports on persistence of viable Mycobacterium avium subsp. paratuberculosis (Map) in pasteurized milk have lend support to the hypothesis that milk consumption can be a cause of human exposure to this pathogen that has been implicated in the etiology of Crohn's disease. The efficacy of high hydrostatic pressure (HHP) for inactivation of micro-organisms has been the object of intensive research, mostly with gram positive and negative cells. However, mycobacteria have never been included in those studies. Here we report an experiment where this method was used to assess the inactivation of Map in milk by moderate HHP treatments held at mild temperatures. Two strains of Map (ATCC 19698 and field isolate 3644/02) were inoculated (4.05x10⁸ CFU/ml and 3.24x10⁸ UFC/ml, respectively) into commercial sterilised milk and submitted to 10 minutes treatments at 300, 400 and 500 MPa at 5 and 20ºC in duplicate. Dilutions of the treated milk where inoculated into duplicate mycobactin J-suplemented agarose solidified Middlebrook 7H9 OADC and Herrold's egg yolk tubes. After 16 weeks of incubation at 37 ºC, colonies were counted and logarithmically transformed to calculate the rate of reduction at each level of treatment. Maximum reduction was 4.8 log at 500 MPa, but culture media and strain also accounted for significant effects. Temperature had no significant effects. HHP milk treatment can be thus considered at least as efficient as pasteurization for Map inactivation, but with the advantage of avoiding the effects of heating on milk properties.