One particular candidate gene, the dopamine D4 receptor (DRD4), has been the focus of intense study regarding ADHD since the original investigation by La Hoste et al, 1 an observation confirmed by a recent metaanalysis. 2 However, two previous studies from Israel failed to observe this association. 3,4 We have now recruited an additional sample and, overall, in the combined sample of 178 triads we observe using the transmission disequilibrium test, 5,6 preferential transmission of the short allele. Additionally, we now report the effect of the DRD4 repeat region on the Test Of Variables of Attention (TOVA), a widely used computerized continuous performance test. 7 Probands with the short exon III repeat performed significantly worse on the TOVA measured both by errors of commission and response time variable. Intriguingly, a 'dose effect' was observed. Increasing repeat size is accompanied by a reduced number of errors of commission and a significant difference is observed between the 2 vs 7 repeats. On the whole, our results lend credence to the notion that the relationship between the DRD4 receptor and ADHD is complex and may be reflecting linkage disequilibrium between the 7 or long DRD4 exon III repeats and a 'true' risk allele in this gene or a neighboring locus. Molecular Psychiatry (2002) 7, 790-794. doi:10.1038/ sj.mp.4001078 In an independently recruited sample of 76 families we again observed preferential transmission (TDT 2 = 3.76, P = 0.05, 1 df) of the short repeat alleles from heterozygous parents to the ADHD proband in the Petak Tikvah cohort (Table 1, new sample). Overall, in the combined Petak Tikvah sample preferential transmission of the short DRD4 repeat was observed (TDT 2 = 9.11, P = 0.0025). Combining all 178 recruited ADHD families (from two geographical locations-Jerusalem and Petak Tikvah), preferential transmission of the short alleles is observed from heterozygote parents to the ADHD proband (TDT 2 = 7.31, P = 0.006).We also analyzed the distribution of the DRD4 exon III polymorphism by case control design comparing