Fused deposition modeling (FDM) is currently one of the most widely utilized prototyping technologies. Studies employing statistical techniques have been conducted to develop empirical relationships between FDM process factors and output variables such as dimensional accuracy, surface roughness, and mechanical properties of the fabricated structures. However, the effects of nozzle temperature (T), nozzle-bed distance (NBD), and their interactions on strut width (SW) have not been investigated. In the present work, a two-way factorial study with three levels of T and NBD in triplicates was undertaken. A fixed-effects model with interaction was proposed and remedial measures based on the error analysis were performed to obtain correct inferences. The factor main/interaction effects were all found to be statistically significant (p < 0.05) using analysis of variance (ANOVA). Multiple comparisons were conducted between treatment means using the Tukey's method. A multiple linear regression (MLR) model (R2 = 0.95) was subsequently developed to enable the prediction of SW. The developed MLR model was verified experimentally; by (1) the fabrication of individual struts and (2) the fabrication of single-layer scaffolds with parallel raster patterns. The percentage error between the predicted and observed widths of individually fabricated struts was 3.2%, and the error between predicted and observed SW/spacing for the single-layer scaffolds was ≤ 5.5%. Results indicate that a similar statistical methodology could be potentially employed to identify levels of T and NBD that yield defined width struts using open architecture, personal or commercial FDM setups, and existing/new materials.