Friedreich ataxia (FA) is an autosomal recessive degenerative disease of the nervous system of unknown biochemical cause. The FA gene has been shown to be in close linkage with the two chromosome 9 markers D9S5 and D9S15, and linkage disequilibrium between FA and D9S15 has been detected in French families by Hanauer et al. We used new highly informative markers at the above loci to analyze Italian FA families for linkage and linkage disequilibrium. The new markers were a three-allele BstXI RFLP at D9S5 (PIC = .55) and a six-allele microsatellite, typed by polymerase chain reaction, at D9S15 (PIC = .75). We obtained maximum lod scores of 8.25 between FA and D9S5, 10.55 between FA and D9S15, and 9.52 between D9S5 and D9S15, all at zero recombination. Our results, combined with those reported by other authors, reduce maxlod-1 (maximum lod score minus 1) confidence limits to less than 1.1 cM between FA and D9S5, 1.2 cM between FA and D9S15, and 1.4 cM between D9S5 and D9S15. Linkage disequilibrium with FA was found only for D9S15 when all families were evaluated but was also found for a D9S5/D9S15 haplotype in a subgroup of southern Italian families. We conclude that FA, D9S5, and D9S15 are tightly clustered and that studies of geographically restricted groups may reveal a limited number of mutations responsible for the disease in the Italian population. We present preliminary evidence from pulsed-field gel electrophoresis that D9S5 and D9S15 may be less than 450 kb apart. Linkage disequilibrium between FA and D9S15 suggests that the disease gene may be at an even shorter distance from this marker locus, which therefore represents a very good starting point for cloning attempts.