In complex genetic disorders, the interval containing a potential susceptibility gene usually is large and cannot be narrowed by searching for linked families with recombinations close to the locus. If linkage disequilibrium can be detected, the interval can be narrowed to 1 cM or less. We had previously obtained evidence for association/linkage disequilibrium by the transmission disequilibrium test (TDT) for the intronic CA repeat marker G-olfα located on chromosome 18p in 59 families with schizophrenia each having at least two affected siblings and their parents (P = 0.00055 for the narrow- and P = 0.000088 for the broad disease definition which included unipolar depression). Since we also detected linkage in this region (Zmax = 2.92 for the broad model, assuming heterogeneity), the sample had to be reduced to independent siblings only, in order to test for linkage disequilibrium in the presence of linkage. The 124 bp allele of G-olfα was transmitted 14 times vs. 2 times nontransmitted (P = 0.0043). The family sample was then combined with a sample of 65 simplex families (triads). The P value remained statistically significant (0.009). There was also a preferential maternal transmission of the allele (P = 0.019). Our results suggest the existence of allelic association/linkage disequilibrium of functional psychoses with the marker G-olfα.