Optimization of protoplast release (202×104 ml−1) from a strain of Rhizoctonia solani anastomosis group three was achieved by treating 3 d old mycelia from broth cultures with Novozyme 234 for 3 h at 30 °C. The only other treatment, of many tested, that yielded any protoplasts (15·6×104 ml−1) was commercial lysing enzymes from Trichoderma harzianum. Various combinations of commercial enzyme preparations, variations of osmotic stabilisers, and mycelial age, were ineffective in raising the protoplast yield. Regeneration of colonies from protoplasts was maximal (83–87%) when protoplasts were incorporated in cooled agar containing 5% glucose, malt extract and yeast extract, supplemented with 0·8 or 1·0 M mannitol as osmotic stabiliser. Liquid culture broth and other potential osmotic protectants at various concentrations yielded fewer regenerating colonies (5% in broth and 13–75% in agar). The Rhizoctonia strain was highly sensitive to hygromycin B, so transformation of protoplasts was achieved with a polyethylene glycol-based procedure, using plasmid pES200 containing the bacterial hygromycin B phosphotransferase gene driven by a trpC promoter from Aspergillus nidulans. Transformation was achieved using the DNase inhibitor aurintricarboxylic acid and incubation temperatures below 10–12 °. It was confirmed by high stringency hybridization of DNA extracted from the transformants, using pES200 as a radiolabelled probe. The transformants grew slowly (25–35 mm diam after 14 d) then ceased growth, and never produced fast-growing sectors indicative of integrated plasmid DNA. Transformation was also achieved with pAXHY2 from Cryphonectria parasitica, and colonies reached 55–65 mm diam before they stopped growing. Southern blots confirmed the presence of non-integrated pAXHY2, but no fast-growing sectors were produced.