The bryophyte vegetation of upland limestone grassland at Buxton in the southern Pennine Hills (UK) was studied following seven years' continuous simulated climate change treatments. The experimental design involved two temperature regimes (ambient, winter warming by 3°C) in factorial combination with three moisture regimes (normal, summer drought, supplemented summer rainfall) and with five replicate blocks. Percentage cover of the bryophytes was estimated visually using 15 randomly positioned quadrats (30 cm × 30 cm) within each of the 30 3 m × 3 m plots. Significant treatment effects were found but these were relatively modest. Total bryophyte cover and cover of Calliergonella cuspidata and Rhytidiadelphus squarrosus responded negatively to drought, whereas Fissidens dubius increased in the droughted plots. Campyliadelphus chrysophyllus increased with winter warming, while R. squarrosus, Lophocolea bidentata and species richness all decreased. The effects on the total bryophyte flora were further studied by canonical correspondence analysis, which yielded a first axis reflecting the combined effects of the moisture and temperature treatments. However, this analysis and a detrended correspondence analysis of the plot data also revealed that natural factors were more important causes of variation in the grassland community than the simulated climate treatments. It was concluded that dewfall may be an important source of moisture for grassland bryophytes and that this factor may have reduced the impact of the moisture treatments. The absence of some thermophilous species such as Homalothecium lutescens in the plots initially may also have reduced their scope for major vegetational change.