Using 10 overlapping nested sets of primers and using peripheral blood lymphocyte (PBL) total RNA as template, we have developed a system, based on PCR, which allows the rapid production of double-stranded cDNA corresponding to the entire coding sequence of the dystrophin gene. The product can be visualized on native minigels by ethidium staining and directly sequenced after gel purification. We have used this system to analyze the structures of PBL dystrophin mRNA in 26 Duchenne, Becker, or intermediate muscular dystrophy patients who have gross rearrangements of the dystrophin gene. In each case, the effect that the genomic rearrangement has on the structure of the transcript--and, by inference, on the dystrophin protein--has been determined, and the results confirm the frameshift hypothesis. The study also identifies a series of alternatively spliced transcripts which are specific to the rearranged genotypes and which seem therefore to arise following the alteration in the context of the splice signal. The system has been used for unambiguous identification of carrier females. Furthermore, the rapid production of microgram quantities of dystrophin cDNA from a readily accessible tissue makes point-mutation screening a practical proposition.