The effects of select system parameters on the performance of direct contact membrane distillation (DCMD), in terms of permeate flux (J) and temperature polarization coefficient (TPC), were investigated and compared for flat-sheet (FS) and hollow-fiber (HF) modules. Three-dimensional (3D) computational fluid dynamics (CFD) models were developed and validated using experimental data. Then, a two-level full factorial design tool was used to design the simulation runs to investigate the effects of four selected process parameters on flux and TPC. These are: feed inlet temperature (T-f), permeate inlet temperature (T-p) and Reynolds number on the feed (Re-f) and permeate (Re-p) sides. The effect of each factor and the interactions thereof were assessed based on data obtained from the CFD models. Although all four parameters showed a significant influence on flux and TPC, their interactions had different effects on the HF and FS modules. For instance, since the permeate stream temperature increases much faster along the permeate channel in the HF module than in the FS module, Re-p has a more substantial impact in the HF module. At higher Re-p values, higher T-f results in further enhancement in permeate flux in the HF module. Furthermore, the CFD model results indicate that the FS module generally showed better performance, in terms of permeate flux, than the HF module under the same DCMD process conditions. In the CFD model runs which yielded the highest flux in both HF and FS modules, the permeate outlet temperatures from the HF and FS modules were 21.72 degrees C and 18.53 degrees C, respectively, under the same operating conditions. Generally, the HF module exhibited about 21% lower flux than the FS module.