This article presents an experimental study of the material extrusion process to examine the impact of less-analyzed factors on surface roughness, not including the commonly known layer height. The study also gives attention to the inner surfaces of printed samples. The design of experiments used a fractional factorial design with five two-level factors: bed temperature, fan percentage, printing speed, printing temperature, and wall thickness. Thus, the purpose of the design is screening. A two-dimensional contact profilometer was used to measure the surface roughness. Generally, the inner faces of the printed samples showed lower surface roughness. In addition, four analysis of variance showed that wall thickness (for Ra and Rz) and printing speed (for Ra) are significant for the outer and inner surfaces, respectively. As the wall thickness increased, the surface roughness decreased. Meanwhile, the lower the printing speed, the worse the surface quality on the inner faces.