Infection caused by M. paratuberculosis has been diagnosed by microbiological isolates, histopathology and immunodiagnosis in wild and captive cervids in many countries. It has emerged as a significant clinical entity in a small proportion of New Zealand's 4000(+) farmed deer herds. The prevalence of infected herds has increased incrementally in the past decade. Unique aspects which characterise Johne's disease (Jd) in farmed deer, is the early onset of clinical disease with significant fatalities in red deer less than 1 year old. Deer can become infected with either 'bovine' or 'ovine' strains of M. paratuberculosis. Not only has cervine Jd become a production limiting disease, it is now a major cause of non-specific mycobacterial sensitisation in deer herds, causing false positive reactions to the tuberculin skin test. This has become a confounding factor in the implementation of deer whole herd testing as part of the National Tb eradication scheme. This paper will consider the changing patterns of sensitisation to Jd, seen from the New Zealand Tb testing database, over recent years. Diagnostic testing within herds with significant levels of Jd infection and clinical disease have been carried out. Patterns of immune reactivity (cellular and humoral) which provide optimal diagnostic precision in naturally infected deer will be outlined. Immune profiles seen in animals vaccinated with live or killed attenuated M. paratuberculosis (316F) vaccines will be discussed. We will attempt to define 'immunological signatures' that are compatible with protection and distinguishable from diagnostic patterns that typify natural infection or disease. Different vaccine formulations are currently being tested using measurements of protective efficacy against experimental infection. Deer may provide an unique model to study Jd infection, pathology, immunopathology and protective immunity, because of the acutely florid presentation of Jd in this group of ruminants.