Despite the widespread use and strong promotion of the sustainable forest management approach, there are still uncertainties about the actual contribution of current forest management practices to sustainability. We studied the problem of sustainable timber production in four tropical countries (Belize, Guyana, Suriname, and Trinidad and Tobago). Data assessed on experimental plots covering 10 km2 were used to compare management practices of four forest tenure types that commonly exist in the study countries: large scale concessions (LSC), private forests (PR), periodic block system forests (PBS), and community managed forests (CM). As an indicator of sustainable timber production, we calculated the recovery times expected under the initial condition of the stands and compared them with currently practiced cutting cycles. Three growth scenarios were simulated using diameter growth rates (1.6/2.7/4.5 mm year−1) from empirical data from studies in the region. Initial volumes were determined for all commercial trees as well as for commercial trees with a DBH-threshold ≥45 cm. Highest initial volumes were found in LSC and PBS managed forests. Lowest volumes were found in CM and PR forests. Assuming the lowest growth rate for all commercial trees, none of the stands studied reached the initial pre-harvest volumes within the currently practiced cutting cycles. Assuming the highest growth rate for all trees, LSC, PBS, and PR forests reach the initial pre-harvest volume. Looking at the subset of commercial trees with a DBH ≥45 cm, all stands will reach the initial volume within 30 years only if the highest growth rate is assumed. We show that general harvest codes do not guarantee sustainable forest management in the tropics. Local stand conditions must always be one of the guiding principles of sustainable timber utilization. Applying the rigid rules, which do not take into account the current conditions of the stands, entails long-term risk of forest degradation.