A total of five Wolf-Hirschhorn syndrome (WHS) patient with a 4p16.3 de novo microdeletion was referred because of genotype -phenotype inconsistencies, first explained as phenotypic variability of the WHS. The actual deletion size was found to be about 12 Mb in three patients, 5 Mb in another one and 20 Mb in the last one, leading us to hypothesize the presence of an extrachromosome segment on the deleted 4p. A der(4)(4qter-p16.1::8p23-pter) chromosome, resulting from an unbalanced de novo translocation was, in fact, detected in four patients and a der(4)(4qter-q32::4p15.3-qter) in the last. Unbalanced t(4;8) translocations were maternal in origin, the rec(4p;4q) was paternal. With the purpose of verifying frequency and specificity of this phenomenon, we investigated yet another group of 20 WHS patients with de novo large deletions (n ¼ 13) or microdeletions (n ¼ 7) and with apparently straightforward genotypephenotype correlations. The rearrangement was paternal in origin, and occurred as a single anomaly in 19 out of 20 patients. In the remaining patient, the deleted chromosome 4 was maternally derived and consisted of a der(4)(4qter-4p16.3::8p23-8pter). In conclusions, we observed that 20% (5/25) of de novo WHS-associated rearrangements were maternal in origin and 80% (20/25) were paternal. All the maternally derived rearrangements were de novo unbalanced t(4;8) translocations and showed specific clinical phenotypes. Paternally derived rearrangements were usually isolated deletions. It can be inferred that a double, cryptic chromosome imbalance is an important factor for phenotypic variability in WHS. It acts either by masking the actual deletion size or by doubling a quantitative change of the genome.