The direct sequencing of an area of DNA using the product of a polymerase chain reaction (PCR) as template is a very powerful technique. It is straightforward and fast and therefore the method of choice in situations such as the analysis of mutations. The reliability of such analysis depends greatly on correct sequencing results. Obtaining the DNA sequence after a PCR amplification and a cloning step contains the risk of sequencing mutations that have been introduced into the DNA by the DNA polymerase upon PCR; the likelihood of such mutations depends on the choice of the enzyme (1 ). Using the product of a PCR reaction directly as template in a sequencing reaction circumvents this problem. DNA molecules containing changes introduced upon PCR never amount to a major portion of the total PCR product even if the mutations occur early in the PCR reaction. This can be illustrated by the following calculation: If one works with chromosomal DNA of an organism with a genome size of 4000 kbp (which is in the range of bacterial genomes) and uses as little as 1 ng of this DNA as template in a PCR reaction, this amount contains approx 230,000 copies of the genome. If a mutation occurs in one molecule in the first cycle of the temperature program, molecules containing this mutation will make up 1/460,000 part of the entire product.