Pairs of large insert clones located in proximity within 384 well storage plates should join contiguous overlapping sets of clones (hereafter contigs) at a predictable frequency. The frequency of clone pairs joining single contigs should be related to distance apart in the plate. Biased frequencies will be caused by errors such as clone contamination, clone duplication, clone insert size variation, and local variations in genome representation. The aim of this study was to detect potential clone contamination errors by biases in the frequencies at which clone pairs that were within the plate, adjacent rows and columns, or neighboring wells joined the same contigs. Simple statistical analyses of three biases in the frequency of pairs of clones coalescing into the same contig were used to identify potential contaminants. Manual editing determined whether the potential contaminants caused gaps or branches within contigs. BAC and YAC contigs of four plants and two animals were examined, including Glycine max (soybean), Arabidopsis thaliana (thrale cress), Oryza sativa (rice), Zea mays (maize), Sorghum bicolor (sorghum), Mus musculus (mouse), and Homo sapiens (human). Clone pairs that formed contigs significantly (0.05 P 0.0001) exceeded expectation with plate pair clones (up to 5-fold, mouse), neighboring clones (up to 3-fold, mouse), and adjacent clones (up to 4-fold, soybean). After manual examination of clone pairs, contigs that contained fewer errors were built.