To mitigate climate change and preserve biodiversity the conservation of tropical forests has to be promoted. Most of the forests in Borneo are permanent forest estates that are selectively logged using conventional logging techniques, causing high damage to the forests. The incorporation of sustainable forest management (SFM) in a REDD+ scheme can help to avert forest degradation. Therefore, it is crucial that above-ground biomass (AGB) values can be operationally estimated. We used Landsat data to develop a crown cover and forest index of lowland dipterocarp forests, which showed a correlation with an R2 of 0.43 to the AGB measurements on the ground. Based on this index a model has been developed, which allowed the estimation of AGB. The accuracy of the model had an R2 of 0.37 with an error of 87.1 t C/ha. Study areas were two forest estates in Sabah: while Tangkulap has been logged conventionally, Deramakot is still managed in accordance with SFM principles. Our research revealed that while Tangkulap lost in average an annual amount of 5.1 t C/ha between 1991 and 2000, Deramakot experienced an average annual recruitment of 5.4 t C/ha between 2000 and 2007. The difference of 10.5 t C/ha/year can be attributed to the different management systems, confirming the hypothesis that SFM in combination with reduced-impact logging helps to protect AGB.