Background: The Borg Rating of Perceived Exertion (RPE) Scale and talk test (TT) are commonly recommended for persons to gauge exercise intensity. It is not known whether they are suitable to estimate substrate partitioning between carbohydrate and fat in persons with SCI. Objective: Investigate substrate partitioning/utilization patterns associated with RPE and TT. Methods: Twelve participants with chronic paraplegia underwent 2 arm crank exercise tests on nonconsecutive days within 2 weeks. Test 1 was a graded exercise test (GXT) to volitional exhaustion. Test 2 was a 15-minute self-selected steady state (SS) voluntary arm exercise bout simulating a brief, yet typical exercise session. Results: For the GXT, very light intensity exercise (RPE < 9) and TT stage before last positive were associated with highest contribution of fat oxidation (∼35%-50%) to total energy expenditure (TEE). Fat oxidation was low at all stages, with the highest rate (0.13 ± 0.07 g/min) occurring at stage 1 (10 W). Corresponding average RPE was 7 ± 2 and the TT was positive for all participants at this stage. For the SS, fuel partitioning throughout exercise was dominated by carbohydrate oxidation (1.47 ± 0.08 g/min), accounting for almost all (∼94%) of TEE with only a minute contribution from fat oxidation (0.02 ± 0.004 g/min). A positive TT was associated with an average contribution of fat oxidation of ∼10%. Conclusions: RPE but not the TT appears suitable to predict exercise intensities associated with the highest levels of fat oxidation. However, such intensities are below authoritative intensity thresholds for cardiorespiratory fitness promotion, and therefore the applicability of such a prediction for exercise prescriptions is likely limited to individuals with low exercise tolerance. © 2012 Thomas Land Publishers, Inc.