A method of detecting a target polynucleotide sequence in a given nucleic acid analyte characterised by the steps of: a. annealing the analyte to a single-stranded probe oligonucleotide A0 to create a first intermediate product which is at least partially double-stranded and in which the 3′ end of A0 forms a double-stranded complex with the analyte target sequence; b. pyrophosphorolysing the first intermediate product with a pyrophosphorolysing enzyme in the 3′-5′ direction from the 3′ end of A0 to create partially digested strand A1 and the analyte; c. (i) annealing A1 to a single-stranded trigger oligonucleotide B and extending the A1 strand in the 5′-3′ direction against B; or (ii) circularising A1 through ligation of its 3′ and 5′ ends; or (iii) ligating the 3′ end of A1 to the 5′ end of a ligation probe oligonucleotide C; in each case to create an oligonucleotide A2; d. priming A2 with at least one single-stranded primer oligonucleotide and creating multiple copies of A2, or a region of A2; and e. detecting a signal derived from the multiple copies and inferring therefrom the presence or absence of the polynucleotide target sequence in the analyte.