The possibility that gene copy number variations play a role in the development of complex disorders is a topic of considerable interest. Recent reports have highlighted the large number of such variations that exist and that their occurrence varies considerably between populations. A recent report has suggested that copy number variations in four genes (GRIK3, EFNA5, AKAP5 and CACNG2) may be associated with schizophrenia. One problem with this area of study is the validation of high throughput methods such as comparative genomic hybridisation, as the latter inevitably generates false positives. We have used two contrasting methodologies to determine the validity of the findings reported above which if true would have major implications for the pathogenesis of schizophrenia. Samples from a UK population were tested using a method of allele quantification by DNA pooling and samples from Belgium and northern Sweden were tested using Multiplex Amplicon Quantification (MAQ). Both methods were used to test DNA samples used in the original investigation. No copy number variations were found for any of the genes in any samples. Our data suggests that more reliable methods need to be used to validate the existence of CNVs before full scale association studies are carried out.