Despite the increasing reliance on breakpoint cluster region (bcr) determinations in diagnosis of chronic myelocytic leukemia (CML), few reports have dealt with the practical aspects of specimen analysis. In the setting of a routine molecular diagnostics laboratory, samples from 68 patients with active CML were evaluated for bcr rearrangements, with the use of a variety of enzymes and two probes. The data have been used to develop an efficient strategy for bcr screening and breakpoint determination. Screening with the universal bcr (UBCR) probe on Xba I and BgI II digests yielded bcr rearrangements in 100% of the Ph1-positive patients and three of the seven Ph1-negative patients, giving bcr analysis a sensitivity of 100%. A single-enzyme screen using the UBCR probe would have resulted in a false negative rate of 10%. The false negative rate was determined during the breakpoint site analysis from additional digests hybridized to both the 3' and UBCR probes. The false negative rate for the 3' probe was 26.5%, because of deletions or 5' rearrangements. The method of breakpoint site determination was dependent on screening results. In 78% of cases, one additional hybridization with two enzyme digests was required. During breakpoint site analysis, a rare false negative result was also demonstrated with Bam HI and Eco RI. This screening strategy has made bcr analysis competitive with cytogenetic analysis at the authors' institution; although turnaround time may be slightly longer, bcr analysis can yield information (such as detecting bcr-positive/Ph1-negative patients and determining breakpoint site) that cannot be obtained by cytogenetics.