Under the Kyoto Protocol industrialized countries will be able to meet carbon emission reduction commitments by financing forestry projects that sequester carbon in developing countries. While this mechanism would compensate for missing markets in forest environmental services, it could also enable industrialized countries to avoid reducing energy use. This paper assesses whether such projects could contribute to improved logging practices in the tropics. Results from studies primarily in Asia and Latin America are analyzed in the context of the modalities of the Kyoto Protocol. Results show that the opportunity cost of shifting from conventional logging to improved practices may have been underestimated. At the same time the long-term carbon and biodiversity benefits of improved forest management may have been underestimated. These results follow primarily from the fact that most previous studies assume that a permanent forest estate is maintained under conventional logging and that cutting cycles are as long as 30-60 years. A more realistic scenario, however, consists of repeated harvesting at short intervals during the first few decades, resulting in the degradation of the forest into shrub and grassland. The implications of these results are that forest management projects may be less cost-effective than previously assumed. Therefore, expectations about their potential contribution to improved management should be scaled down. At the same time, the extent to which such projects will enable industrialized countries to avoid reducing industrial pollution is also unlikely to be significant. Cost-effectiveness is likely to be highest where timber volumes in the first few decades after initial logging are comparable under conventional and improved logging. This is likely where topography is relatively flat, biodiversity values are low, wastage of felled timber is high and the policy environment is favorable. A number of proactive measures are suggested to expand the niche for forest management carbon projects. These measures are justified because the incremental carbon and biodiversity benefits in the long run may be higher than previous studies have indicated.