Tropical forests contain a considerable fraction of the global soil carbon (C) stock, but the response of tropical soil C to predicted changes in primary productivity remains poorly understood. Drastic changes in soil C storage and loss are likely to occur if global change alters plant net primary production (NPP). This research assessed the effects of a decade of litter removal and addition on soil C storage in a seasonal tropical forest in Panama. Previous manipulation studies like the one used here have reported strong responses of carbon pools with addition and removal of litter in tropical forest ecosystems. Specifically, total C gains have been observed after addition while corresponding losses of C have been found with litter removal. However, past manipulation experiments fail to account for the heterogeneous nature of carbon cycling in these systems. The overarching hypothesis addressed here, is that changes in net primary production can affect the storage of soil C on various timescales. To address this, a density fractionation scheme was used to assess manipulation effects on rapidly- and slowly-cycling pools of C. Soil samples were collected from 0-5 cm depth in 15 - 45 x 45m plots with litter removal (L-), 2x litter addition (L+), and control (n=5). To account for any potential fertilization effect from added litter, soils from a long-term fertilization experiment were also used. Soil C storage in all pools was significantly different among the three litter treatments, and some seasonality resulted in every pool. L+ increased the oldest, most stable pool of C by 15% in the dry season and 28% during the wet season. Alternatively, L- decreased C concentrations in this long-term pool by 26% in the dry season and 28% in the wet season relative to control. There was no indication of any fertilization effect on the storage of C in this forest. The results here demonstrate that changes in primary productivity have the potential to alter the storage and cycling of C in lowland tropical forest soils.