Repair of DNA double-strand breaks (DSB) is essential for cell viability and genome stability. Homologous recombination repair plays an important role in DSB repair and impairment of this repair mechanism may lead to loss of genomic integrity, which is one of the hallmarks of cancer. Recent research has shown that the tumor suppressor genes p53 and BRCA1 and -2 are involved in the proper control of homologous recombination, suggesting a role of this type of repair in human cancer. We developed a novel assay based on recombination between two Green Fluorescent Protein (GFP) sequences in transiently transfected plasmid DNA. The plasmid construct contains an intact, emission-shifted, "blue" variant of GFP (BFP), with a 300 nucleotide stretch of homology to a nonfunctional copy of GFP. In the absence of homologous recombination only BFP is present, but homologous recombination can create a functional GFP. The homologous regions in the plasmid were constructed in both the direct and the inverted orientation of transcription to detect possible differences in the recombination mechanisms involved. A panel of human tumor cell lines was chosen on the basis of genetic background and chromosome integrity and tested for homologous recombination using this assay. The panel included cell lines with varying levels of karyotypic abnormalities, isogenic cell lines with normal and mutant p53, isogenic cell lines with or without DNA mismatch repair, BRCA1 and -2 mutant cell lines, and the lymphoma cell line DT40. With this assay, the observed differences between cell lines with the lowest and highest levels of recombination were about 100-fold. Increased levels of recombination were associated with mutant p53, whereas a low level of recombination was found in the BRCA1 mutant cell line. In the cell line HT1080TG, a mutagenized derivative of HT1080 with two mutant alleles of p53, high levels of recombination were found with the direct orientation but not with the inverted orientation plasmid. No difference in recombination was detected between two isogenic cell lines that only differed in DNA mismatch repair capability. We conclude that this assay can detect differences in homologous recombination capacity in cultured cell lines and that these differences follow the patterns that would be expected from the different genotypes of these cell lines. Future application in normal cells may be useful to identify genetic determinants controlling genomic integrity or to detect differences in DNA repair capacity in individuals.