OBJECTIVES: This study screened the design parameters of a reduced-diameter implant to determine which parameters have the most significant effect on the implant fatigue limit. METHODS: A dental implant assembly, which included an implant body (Biomet 3i), an abutment (GingiHue®), and an abutment screw (Gold-Tite Square screw) was scanned using micro-computed tomography (SkyScan 1172) and was measured using Mimics (Materialise) and an optical microscope (VHX-1000, Keyence). Sixteen design parameters were measured, and the values of the commercial design were taken as reference level for each design parameter. Values up to 20 % lower and 20 % higher than the reference were explored using a Taguchi orthogonal array (DOE++, Reliasoft), which varies more than one design parameter at a time to efficiently explore all main effects and lower order interactions across few implant designs. Solid models of these 27 implant designs and the reference design were constructed using SOLIDWORKS (Dassault Systèmes). Each solid model was loaded according to ISO 14801. fe-safe (Dassault Systèmes) was used to estimate the fatigue limits. ANOVA statistical test in DOE++ was used to screen the design parameters. RESULTS: Interaction between the coronal and apical tapers of the implant body had a significant effect on the fatigue limit (p ≤ 0.05), where fatigue limit was low for designs with a constant taper. Conversely, the combination of high degree of apical taper and low degree of coronal taper lead to the highest fatigue limit. SIGNIFICANCE: Using a Taguchi orthogonal array proved to be an efficient strategy for screening implant design parameters for effect on fatigue limit. The modified implant designed by manipulating the most influential parameters is predicted to have much greater fatigue limit compared to the commercially available design.