This study investigates (a) the extent to which L2 learners use phonetic information from visual cues to improve the perception of a novel phonemic contrast following intensive perceptual training, and (b) the impact of perceptual training on pronunciation. 62 Japanese learners of English were tested on their perception of the /l/-/r/ contrast in audio, visual and audiovisual modalities, and then undertook ten sessions of perceptual training before being tested again. Eighteen were trained using auditory stimuli, 25 using natural audiovisual stimuli and 19 using audiovisual stimuli with a synthetic face. /l/-/r/ perception improved in all groups but learners trained audiovisually did not improve more than those trained auditorily. Auditory perception improved most for 'A training' learners and sensitivity to visual cues improved most for 'natural AV training' learners. The learners' pronunciation of /l/-/r/ improved significantly following perceptual training, with a greater improvement seen for those trained audiovisually with natural stimuli.