A method for determination of zygosity. Two sets of primers are used. Of the inner set, primer A is complementary to a point mutation on the coding strand and primer B is complementary to the wild-type non-coding strand. The outer set of primers, P and Q, flank the mutation site on the coding and non-coding strands, respectively. P and Q are chosen so that the amplicons PB and AQ will have characteristic lengths. The pattern of PCR products is informative of the zygosity of the tested genomic DNA: all samples will generate PQ; the heterozygote and the homozygous wild-type will generate PB and the heterozygote and homozygous mutant will generate AQ. A and B are designed to mismatch the unintended allele at their 3' ends and they possess G+C-rich tails.

The efficiency of replication is affected by template transfer: as replication of an amplicon proceeds, it replaces genomic DNA as the template. In bi-PASA, as PQ accumulates, it also can serve as a template for amplification of the shorter fragments. This is contrasted with self amplification, which occurs when the shorter templates replicate only themselves. Self amplification is favoured by the G+C-rich tails and by annealing conditions that discourage all but the stronger G-C bonds. Basically similar to bi-PASA is tetra-primer PCR, in which the internal set of primers are mismatched in the middle of their sequences, and they lack G+C-rich tails. This variation requires high- and low-stringency annealing conditions to generate appropriate amounts of the three potential amplicons, whereas bi-PASA uses a constant annealing temperature.