We evaluated various aspects of grand rounds videoconferenced from a tertiary care hospital to a regional hospital in Nova Scotia. During a five-month study period, 29 rounds were broadcast (19 in medicine and 10 in cardiology). The total recorded attendance at the remote site was 103, comprising 70 specialists, nine family physicians and 24 other health-care professionals. We received 55 evaluations, a response rate of 53%. On a five-point Likert scale (on which higher scores indicated better quality), mean ratings by remote-site participants of the technical quality of the videoconference were 3.0-3.5, with the lowest ratings being for ability to hear the discussion (3.0) and to see visual aids (3.1). Mean ratings for content, presentation, discussion and educational value were 3.8 or higher. Of the 49 physicians who presented the rounds, we received evaluations from 41, a response rate of 84%. The presenters rated all aspects of the videoconference and interaction with remote sites at 3.8 or lower. The lowest ratings were for ability to see the remote sites (3.0) and the usefulness of the discussion (3.4). We received 278 evaluations from participants at the presenting site, an estimated response rate of about 55%. The results indicated no adverse opinions of the effect of videoconferencing (mean scores 3.1-3.3). The estimated costs of videoconferencing one grand round to one site and four sites were C$723 and C$1515, respectively. The study confirmed that videoconferenced rounds can provide satisfactory continuing medical education to community specialists, which is an especially important consideration as maintenance of certification becomes mandatory.