Secondary succession is changing the character of many temperate forests and often leads to closed-canopy stands. In such forests set aside for conservation, habitat management alternatives need to be tested experimentally, but this is rarely done. The Swedish Oak Project compares two often debated alternatives: minimal intervention and non-traditional active management (conservation thinning) on plots of each type replicated at 25 sites. We study responses of several taxa, and here report results for land molluscs. They are considered to be sensitive to more open, drier forest and we predicted a negative effect of the thinning (26% reduction of the basal area; mean value for 25 experimental forests). We sampled molluscs in the litter in ten 20 x 25 cm subplots, and by standardised visual search, in each plot. In total, we recorded 53 species of snails and slugs (24 369 individuals) and the mean species richness in plots was 17. Two seasons after thinning, mean (± SE) species richness had decreased by 1.4 (± 0.9) species in thinning plots, but increased by 0.7 (± 1.0) species in minimal intervention plots, a significant but small change with considerable variation among sites. In matched comparisons with minimal intervention, thinning reduced the overall abundance of molluscs. Most species responded negatively to thinning - but only five of the 53 species were significantly affected, and reproduction seemed to be negatively affected in only one species. An ordination analysis did not reveal any particular change in the species community due to thinning. Thus, the negative effect of conservation thinning on land molluscs was apparently mild - one reason was that many trees, shrubs and other forest structures remained after the treatment. Conservation thinning may be recommended, since other taxa are favoured, but minimal intervention is also a useful form of management for molluscs and saproxylic taxa.