Ninety-two Mycobacterium bovis isolates from cattle, deer and badgers in Northern Ireland and the Republic of Ireland were genotyped by spacer-oligotyping (spoligotyping) and 67 of these were analysed by restriction fragment length polymorphism (RFLP). RFLP analysis was performed using three DNA probes, PGRS, DR and IS6110. Forty-seven of the M. bovis isolates were from 45 different sources; these were typed using both RFLP and spoligotyping. These 47 isolates could be differentiated into 24 different RFLP types and 15 distinct spoligotypes. Although RFLP was found to be more discriminatory compared to the present spoligotyping technique, spoligotyping was able to differentiate 21 RFLP type 'ACA' isolates into three different patterns. The remaining 45 M. bovis isolates were from a small case study, involving infected cattle, deer and badgers from the same geographic region. All these isolates were analysed by spoligotyping and a selection of 20 isolates were RFLP typed. All the isolates in the case study had the same spoligotype pattern with the exception of one cervine isolate. Similarly all the isolates typed by RFLP had the same pattern. Consequently, the predominant strain in the case study was not host restricted. The consistency between the results obtained using the two techniques indicates the potential value of both techniques for epidemiological studies. Spoligotyping was found to be a much more rapid technique and easier to perform, requiring less sophisticated computer software for strain typing. Spoligotyping results were more readily documented and analysed and the technique was also more suitable than RFLP analysis for large-scale screening studies.