Crohn's disease is an idiopathic chronic, panenteric, intestinal inflammatory disease. Its etiology may be infectious. Putative organisms implicated are the measles virus and Mycobacterium paratuberculosis (M. ptb). Empirical anti-mycobacterial treatment is not efficacious in the therapy of Crohn's disease. We suggest that within the spectra of Crohn's disease there are two clinical manifestations, analogous to the tuberculoid and lepromatous manifestations of leprosy. One is aggressive and fistulizing (perforating) the other is contained, indolent and obstructive (non-perforating) (PNAS 91;12721-12724; 1994). Surgical ileal mucosal specimens were from both forms of Crohn's disease as well as controls (n=4 each group). Total RNA was reverse transcribed to cDNA using random primers. RNA was quantified by PCR amplification using published or custom designed primers. Evaluated were; a house keeping gene ß-Actin; a M. ptb specific DNA sequence (IS 900); the cytokine IL-1ß (previously shown to differ between the two forms of Crohn's and used to show lack of contamination). Following PAGE and autoradiography, bands were excised, DNA purified, subcloned and sequenced. Controls included M. ptb (ATCC). All the cases of Crohn's disease and both samples from the patients with ulcerative colitis showed a band comigrating with the M. ptb DNA. There was no signal in the cancer patients. DNA sequence analysis (to date performed on two samples) confirms that this is M. ptb. The IL-1ß data show that these esults cannot be ascribed to sample contamination. The finding of M. ptb DNA in Crohn's disease has been previously reported. However, a cause and effect relationship has not been established. In part, this is due to the lack of efficacy with empirical anti-mycobacterial therapy. We conclude that clinical trials with anti M. ptb therapy are indicated in Crohn's disease. However, patients should be stratified into the aggressive (perforating) or contained forms prior to randomization.