Southeast Asian peatlands have undergone recent land use change with an increase in industrial agricultural plantations, including oil palm. Cultivating peatlands requires creating drainage ditches and other surface microforms (i.e., harvest paths, frond piles, cover plants, and next to the palm). However, it is currently unclear how these management actions affect rates of carbon losses from the peat. Here we report carbon fluxes from each of the different surface microforms measured monthly (soil CO 2 [total soil respiration-R tot ] and stem CH 4) and bimonthly (soil CH 4 , drain CO 2 and drain CH 4). We calculated annual carbon fluxes and partitioned heterotrophic (R h) and root-rhizosphere respiration by sampling rhizosphere and root-free soil. Linear mixed effect models were used to determine which environmental factors best-predicted carbon fluxes, and to develop recommendations for management solutions that could reduce carbon losses. Carbon fluxes varied significantly between the different microforms; the greatest CO 2 fluxes were measured next to the palm and the greatest CH 4 fluxes were measured from the drainage ditches. Annual estimates of R tot , R h and drain CO 2 were 22.08 ± 0.50, 17.75 ± 1.54, and 1.5 ± 0.10 Mg CO 2-C ha −1 yr −1 , respectively. R h varied between the two plantations: Sebungan averaged 11.43 ± 1.37 Mg CO 2-C ha −1 yr −1 and Sabaju averaged 24.08 ± 1.42 Mg CO 2-C ha −1 yr −1. Net ecosystem CH 4 fluxes averaged 61.02 ± 17.78 kg CH 4-C ha −1 yr −1-similar to unmanaged swamp forests. The two plantations did not vary in overall CH 4 flux, but did vary in transport pathway. CH 4 fluxes from the soil, drains and stems followed a ratio of 50:50:0 from Sabaju (water table depth [WTD]: −0.49 ± 0.004 m) and 11:98:0 from Sebungan (WTD: −0.77 ± 0.007 m). R h dominated the peat carbon losses. WTD controlled variation in R h from Sebungan where the WTD was deeper. Air and soil temperature controlled variation in Sabaju, with greater fluxes from the harvest path, attributed to the absence of shade. These results suggest that shading the soil (e.g., through addition of frond piles) and raising the water table may be the most effective ways to reduce peat carbon loss from drained peat soils.