Individuals with the 18q- syndrome have variable deletions from the long arm of chromosome 18. They also exhibit a highly variable phenotype. To correlate genotype with phenotype accurately, extensive molecular and phenotypic analyses are needed on each affected individual. As a part of this analysis, we have determined the parental origin of the deleted chromosome in 34 individuals with the 18q- syndrome. We have found that 85% of the de novo deletions are paternal in origin. The percentage of fathers of individuals with paternally derived deletions who were > 30 years old was (not significantly) greater than that of the general population. The mothers of individuals with maternally derived deletions were near an average age for childbearing compared to the general population. Individuals with maternally derived terminal deletions had breakpoints as varied as those with paternally derived deletions. These results are consistent with the hypothesis that the reduced incidence of maternally derived deletions is not due to reduced viability, since individuals with large maternally derived deletions of chromosome 18q were found. We hypothesize that the prevalence of paternally derived deletions is due to an increased frequency of chromosome breakage in male germ cells. These results are consistent with results observed in other segmental aneusomies in which there is a high incidence of paternally derived deletions.