BackgroundComparative genomic hybridization (CGH) is a powerful method to investigate chromosomal imbalances in tumor cells. However, DNA quantity and quality can be limiting factors for successful CGH analysis. The aim of this study was to investigate the applicability of degenerate oligonucleotide‐primed PCR (DOP‐PCR) and a recently developed linker‐adapter‐mediated PCR (LA‐PCR) for whole genome amplification for use in CGH, especially for difficult source material.MethodsWe comparatively analyzed DNA of variable quality derived from different cell/tissue types. Additionally, dilution experiments down to the DNA content of a single cell were performed. FISH and/or classical cytogenetic analyses were used as controls.ResultsIn the case of high quality DNA samples, both methods were equally suitable for CGH. When analyzing very small amounts of these DNA samples (equivalent to one or a few human diploid cells), DOP‐PCR‐CGH, but not LA‐PCR‐CGH, frequently produced false‐positive signals (e.g., gains in 1p and 16p, and losses in chromosome 4q). In case of formalin‐fixed paraffin‐embedded tissues, success rates by LA‐PCR‐CGH were significantly higher as compared to DOP‐PCR‐CGH. DNA of minor quality frequently could be analyzed correctly by LA‐PCR‐CGH, but was prone to give false‐positive and/or false‐negative results by DOP‐PCR‐CGH.ConclusionsLA‐PCR is superior to DOP‐PCR for amplification of DNA for CGH analysis, especially in the case of very limited or partly degraded source material. © 2004 Wiley‐Liss, Inc.