We used a pre-post design to compare the health and care utilization of patients receiving telehaemodialysis services in two James Bay Cree communities. The Cree are an Amerindian First Nation living in the remote James Bay region. The same group of dialysed patients (n = 19) was followed longitudinally over a two-year period: 12 months pre and 12 months post. Analysis of variables measuring the patients' health conditions showed that the quality of care provided was well within recognized good practice guidelines. Repeated measures ANOVA on the variables measuring care utilization showed a significant decrease in the monthly number of medication changes over time (P < 0.01). Different telehaemodialysis models were used in the two communities (virtual patient rounds and telecase reviews with multidisciplinary teams), but they did not lead to differences in health condition or care utilization. This suggests that there is no single prescriptive model for the delivery of tele-expertise.