Experiments were carried out to improve shootformation from leaves of wild cherry (Prunusavium L). Emphasis was placed on maximising shootnumbers per leaf. Results with one genotype showedthat leaf size was critical in determining whethershoots formed, with leaves 3–5 mm in length formingmost shoots. Leaves larger than this failed to formcallus or shoots, while smaller leaves tended todevelop into fast-growing callus. There weresignificant differences among ten genotypes, selectedarbitrarily, in proportions of leaves thatregenerated, numbers of shoots formed and numbers ofpositions per leaf where shoots formed. Supplementingmedium with the surfactants Tween-20, and PluronicsF127 and F68 failed to increase the proportions ofleaves of the one genotype tested that producedshoots. However, Tween-20 increased numbers of shootsper leaf and the numbers of positions per leaf thatformed shoots. The main effects of Pluronics F68 andF127 with increasing concentration over the rangetested (10–1000 mgl−1) were significantdecreases in shoot numbers and numbers of positionsper leaf where shoots formed. Using the best genotype,accession 1905, leaves 3–5 mm in length, and Tween-20at 10 mgl−1, 84% (s.e. ± 2.3%) of leavesformed a mean (± s.e.) of 3.5 ± 0.3 shootsfrom 2.0 ± 0.2 locations per leaf.