Adoption of reduced-impact logging (RIL) methods could reduce CO2 emissions by 30-50% across at least 20% of remaining tropical forests. We developed two cost effective and robust indices for comparing the climate benefits (reduced CO2 emissions) due to RIL. The indices correct for variability in the volume of commercial timber among concessions. We determined that a correction for variability in terrain slope was not needed. We found that concessions certified by the Forest Stewardship Council (FSC, N=3), when compared with non-certified concessions (N=6), did not have lower overall CO2 emissions from logging activity (felling, skidding, and hauling). On the other hand, FSC certified concessions did have lower emissions from one type of logging impact (skidding), and we found evidence of a range of improved practices using other field metrics. One explanation for these results may be that FSC criteria and indicators, and associated RIL practices, were not designed to achieve overall emissions reductions. Also, commonly used field metrics are not reliable proxies for overall logging emissions performance. Further, the simple distinction between certified and non-certified concessions does not fully represent the complex history of investments in improved logging practices. To clarify the relationship between RIL and emissions reductions, we propose the more explicit term “RIL-C” to refer to the sub-set of RIL practices that can be defined by quantified thresholds and that result in measurable emissions reductions. If tropical forest certification is to be linked with CO2 emissions reductions, certification standards need to explicitly require RIL-C practices.