Amyotrophic lateral sclerosis (ALS) is a progressive motor neurodegeneration resulting in paralysis and death from respiratory failure within 3-5 years. About 20% of familial cases are associated with mutations in the gene for copper/zinc superoxide dismutase ( SOD1 ), which catalyses the dismutation of the superoxide radical to hydrogen peroxide and oxygen. Experimental evidence suggests mutations act by a toxic gain of function but the mechanism is unknown. There are >60 known SOD1 mutations associated with ALS and all are dominant except for one in exon 4, a D90A substitution which is recessive. D90A pedigrees with dominant inheritance have now been reported and this apparent contradiction needs to be explained. We performed a worldwide haplotype study on 28 D90A pedigrees using six highly polymorphic microsatellite markers. We now show that all 20 recessive families share the same founder (alpha = 0.999), regardless of geographical location, whereas several founders exist for the eight dominant families (alpha = 0.385). This finding confirms that D90A can act in a dominant fashion in keeping with all other SOD1 mutations, but that on one occasion, a new instance of this mutation has been recessive. We propose a tightly linked protective factor which modifies the toxic effect of mutant SOD1 in recessive families.