1. Acid-base status has major effects on diatoms, but there is little information on their shortterm response to changing acidity. This is despite the use of diatoms as bioindicators in streams where acid episodes are important during rainstorms (hours to days) or snowmelt (days to weeks). In the Llyn Brianne experimental catchments (Wales, UK), we attempted to mimic the effects of short-term acidification by (i) reciprocally transplanting diatoms between two streams of contrasting acidity and (ii) using acid-diffusing substrates. 2. Diatom diversity decreased rapidly on substrata transplanted from the circumneutral into the acidic stream, and increased in the reciprocal transplantation. Changes in dominant taxa occurred within three days in the acidic stream because of the rapid growth of Eunotia exigua, and by nine days in the circumneutral stream because of the proliferation of Achnanthidium minutissimum. Transplants were near indistinguishable from ambient assemblages by day 12. 3. There were no effects of enclosures on assemblage composition, but diatoms responded more rapidly to altered chemistry in enclosures with coarse mesh (26 · 50 mm) than finer mesh (320 lm). 4. Chemical diffusing substrates comprised terracotta tiles attached to dosing reservoirs that created locally acid (using H 2 SO 4 ) or metal-rich conditions (using MnSO 4 ) in the circumneutral stream over 26 days. Diatom responses were compared with reference substrates dosed with deionised or circumneutral stream water, and we also assessed whether effects were moderated by macroinvertebrate grazers. 5. Surface pH was lower by 1-2 pH units on acid-dosed substrates than on reference tiles or in surrounding streamwater. Grazed assemblages on acid-dosed substrates differed significantly from ungrazed reference assemblages, acquiring significantly greater relative abundance of Eunotia spp. However, the magnitude of response was less than in the between-stream transplantations either because (i) metal exposure and base cation concentrations differed between the transplants and dosing substrates or (ii) diatom response to reduced pH on the diffusing substrates was restricted by the scarcity of acidobiontic diatoms in the circumneutral stream. Similar filter, founder or dominance effects might also affect diatom responses to real acid episodes.6. These data show that diatom assemblages can respond rapidly and directly to changes in acid-base status, but short-term acidification might affect diatoms more rapidly than subsequent recovery. Because the experimental methods used were imperfect representations of episodic effects, diatom response to real acid events requires further field evaluation.