The growing insight into genetically caused diseases induced by single nucleotide polymorphisms (SNPs) calls for the development of analytical techniques to detect these SNPs. The techniques should be quick, reliable, and inexpensive. In recent years several strategies for genetic analyses were introduced, which are based on polymerase chain reactions (PCR), [1] enzymatic digestion, [2] electrochemistry, [3] or methods where target binding directly changes the fluorescence. [4] We now present a new concept for mismatch detection using chemical DNA cleavage on a solid support (Scheme 1): A single DNA strand, which carries a cleavage site at the central position and a detection tag at the 5'-end, is immobilized on the solid support through the 3'-end and hybridized with the target strand. Upon irradiation the immobilized single strand is cleaved and releases the tag-containing fragment into the solution if a mismatch is present. In the matched case, the higher melting point prevents dehybridization.As the cleavage site we used the 5'-o-nitrophenyl thymidine (1, T*; Scheme 2), which selectively recognizes adenosine and quantitatively cleaves the DNA strand upon irradiation (360 nm) in solution, whereby the whole DNA strand is cleaved.[5] This modified nucleotide 1 was incorporated through standard phosphoramidite chemistry into the modified 21mer 2, which was immobilized on agarose (formation of 3; Scheme 3).[6] We used dispers red 1 (DR1)as a tag, [7] and directly observed the color changes of the beads after irradiation (10 8C, 10 min, l = 360 nm). The results were quantified by HPLC analysis of the tagged fragment 4, which was released into the solutions by irradiation. Subsequent heating of the solid-phase system quantitatively released the remaining tag-containing fragment 4 into solution (Scheme 3, Experimental Section). [8] Irradiation of the immobilized single-strand 3 released 98 % of the cleavage product 4 into the solvent, and the beads became colorless. Thus, a quantitative cleavage of the modified strand occurs not only in solution [5] but also if the strand is immobilized on agarose. This result is in accord with other photolysis experiments with agarose as a solid support.[9] Double-strand formation was quantitative when the immobilized strands 3 were hybridized with a slight excess of target strands, even if one or two mismatches were introduced (Experimental Section). But in contrast to the single-strand experiment, irradiation of a matched double strand at 10 8C released only a small amount of 10mer 4 into the solution (15 %), and the beads remained red. Subsequent heating quantitatively released the cleaved strand 4 (the beads became colorless).[10] This behavior changed if mismatches were incorporated. Irradiation of double strands with two mismatches released 70-98 % of the 10mer 4 into solution at 10 8C (Table 1).The melting points of the photocleaved systems are decreased by two mismatches to such an extent that the tagcontaining fragment 4 is released into the solvent. Experiments in homogen...