Study design: Randomized controlled trial of exercise training in persons with spinal cord injury. Objective: The purpose of this study was to examine the e ects of 9 months of twice-weekly exercise training on strength, arm ergometry performance, and indices of psychological wellbeing and quality of life. Setting: Centre for Health Promotion and Rehabilitation, McMaster University, Hamilton, Ontario, Canada. Methods: Thirty-four men and women (aged 19 ± 65 years) with traumatic spinal cord injury (C4 ± L1; ASIA A ± D) of 1 ± 24 years duration volunteered to participate, and were randomized into exercise (EX; n=21) and control (CON; n=13) groups. Twenty-three subjects (11 EX; 12 CON) successfully completed the 9-month study. Subjects were assessed for one repetition maximum (1RM) strength, arm ergometry performance, and several indices of quality of life and psychological well-being at baseline, 3, 6, and 9 months. Results: At baseline, there were no signi®cant di erences between groups in age, submaximal arm ergometry performance, muscle strength, or psychological well-being. Following training, the EX group had signi®cant increases in submaximal arm ergometry power output (81%; P50.05), and signi®cant increases in upper body muscle strength (19 ± 34%; P50.05); no signi®cant changes occurred in CON. Participants in EX reported signi®cantly less pain, stress and depression after training, and scored higher than CON in indices of satisfaction with physical function, level of perceived health and overall quality of life (P50.05). Exercise adherence (per cent of prescribed sessions attended) in those subjects who completed the 9 months of training was 82.5%.Conclusions: These results demonstrate that long-term twice-weekly exercise training in this population is feasible, and results in signi®cant gains in both physical and psychological wellbeing.