Culture conditions, and other variables that modulate a cell's physiology, can bias a polymerase chain reaction (PCR) amplification against generating a representative population profile. Two Pseudomonas putida nahR alleles were constructed in pUC19 that differ solely in a 31-bp internal segment whose sequence has been inverted. After PCR amplification, the products could be distinguished on the basis of a change in a unique restriction site. When an Escherichia coli strain carrying one nahR allele is submitted to different growth conditions, the consequences of such variations on the relative PCR amplification of whole cells can be ascertained through coamplification with a strain carrying the other allele and subsequent restriction analysis. Cells in stationary phase displayed improved amplifiability while cells grown at 42 degrees C were equally amplifiable as compared to cells grown at 37 degrees C. However, sublethal levels of tetracycline or growth in minimal medium made the PCR target in these cells relatively less amplifiable. When cells are completely lysed and the plasmid DNA is purified beforehand, the coamplification bias is eliminated. These results suggest that mixed populations containing cells in different physiological states may not be representatively amplified by PCR unless a DNA extraction step is included.