Mutations in the cystatin B (CSTB) gene underlie progressive myoclonus epilepsy of Unverricht-Lundborg type (EPM1) (Pennacchio et al., 1996). We previously described an unstable minisatellite expansion mutation in the putative promoter region of CSTB that accounts for the majority of EPM1 patients. Sequencing of a genomic lambda clone, generated from a Finnish EPM1 patient homozygous for an enlarged restriction fragment, revealed a 15- to 18-mer minisatellite repeat expansion (Virtaneva et al., 1997). Later, sequencing of plasmid clones generated from Swiss and French patients revealed a dodecamer repeat expansion (Lalioti et al., 1997a). By restriction enzyme analysis of our original patient clone and a clone generated from an Italian patient, we now show that the expansion is neither a 15-mer nor an 18-mer contrary to our initial results. Moreover, direct sequencing of the Finnish patient clone with Pfu exo polymerase confirmed that the expanded repeat is a dodecamer. Based on this finding and additional experiments, we suggest that the discrepancy between the two studies was due to errors caused by the combination of native Pfu polymerase and modified guanosine deaza-7-dGTP used in the PCR reaction.