The ability to inactivate specific genes in pathogenic organisms is an important requirement for determining the role of those genes in pathogenic processes and for producing attenuated strains as live vaccine candidates. We have investigated the usefulness of two homologous recombination techniques, for producing allelic exchange mutants of Mycobacterium paratuberculosis. In both techniques, a suicide plasmid construct is made containing a DNA fragment encoding a selected gene that is interrupted by insertion of a hygromycin resistance gene. In one technique, the suicide plasmid is delivered into a virulent strain of M. paratuberculosis by high-temperature electroporation, while in the other technique it is delivered in a temperature-sensitive phage construct. The transformed strains are then cultured on solid hygromycin-containing medium. Mutants that are antibiotic resistant are sub-cultured and characterized by PCR and Southern blotting to determine if allelic exchange of the active gene for the inactive gene has occurred. Two virulent M. paratuberculosis strains were used for these studies; strain 989, a New Zealand strain isolated from cattle, and strain k10, the strain used for genome sequencing. When a direct comparison of the electroporation technique to the phage technique was made between these two strains for allelic exchange of two unrelated genes, the phage technique was found to be much more efficient. Efficiency, defined as the percentage of allelic exchange mutants to the number of antibiotic resistant colonies, ranged from 25% to 91 % for the phage technique, and the efficiencies were moderately higher in strain 989 than in strain k10. In comparison, on the one occasion that electroporation was successful, the efficiency was only 10%. Clearly, the phage technique is the preferably approach for allelic exchange in M. paratuberculosis.