Physicians in rural communities have limited access to continuing medical education (CME) opportunities. We hypothesized that CME could be delivered via a telemedicine network as effectively as in-person. Our institution delivers CME lectures and grand rounds in Burlington, Vermont, for in-person attendees, and also via a telemedicine network that links 14 hospitals in Vermont and rural northeastern New York. All participants complete an evaluation questionnaire to receive CME credit. We compared the questionnaire responses of those attending in person with those attending via the telemedicine network. From October 1, 2000 to June 30, 2003, there were 4733 CME sessions, 650 of which had both in-person and telemedicine attendees. Responses from these 650 sessions were compared. Most questions relating to lecture quality scored higher for in-person attendees. Compared to having the presenter in the room, telemedicine attendance was judged to be "more effective" in 19% (n = 334), "as effective" in 60%, (n = 1074), and "less effective" in 21% (n = 367). Eighteen percent of telemedicine attendees said they would have traveled to attend the session. Telemedicine-delivered CME was considered at least as effective as in-person CME 79% of the time. Travel was avoided for 18% of the remote attendees. CME was delivered where it would not have been obtained for 82% of the remote attendees. Telemedicine systems can be used to deliver CME, in spite of lower overall ratings compared to in-person attendance.