Highly uniform and large-area single-walled carbon-nanotube (SWNT) networks are realized by the dip-coating method, which is based on fundamental fluid-dynamic phenomena such as capillary condensation and surface tension. The changes in the polarity and hydration properties of the substrate affect the morphology of the SWNT networks and result in nonlinear growth of the networks in the repetitive dip-coating process. The density and the thickness of the SWNT networks are controlled by processing variables including number of dip coatings, concentration of SWNT colloidal solution, and withdrawal velocity. The networks have uniform sheet resistances and high optical transmittance in the visible wavelength range.