The main purpose of this research is to bolster mechanical properties of the parts, produced by an extrusion-based 3D printer, or fused deposition modeling machine, via increasing the content of continuous fiber yarn to its practical limit. In-melt continuous glass fiber yarn embedding was applied as a reliable and consistent method for simultaneous impregnation to produce continuous fiber-reinforced thermoplastic composites in the fused deposition modeling process. It is well known that the main weakness in the fused deposition modeling 3D printed products is their low strength compared to the manufactured ones by conventional methods such as injection molding and machining processes. This characteristic can be related to both inherent weakness of thermoplastic materials and poor adhesion between the deposited rasters and the layers. Although various attempts have been performed to tackle this issue, it is widely believed that using continuous fibers is the most effective method to serve this purpose if a reliable and consistent method is implemented. Since the mechanical properties of continuous fiber-reinforced composites directly depend on the content of fiber volume, maximizing the fiber content as well as producing an integrated part was assumed as the main objective. In this work, an analysis of various patterns of raster deposition was conducted, followed by the experiments and verification. The effective parameters on the fiber yarn volume, such as fiber yarn diameter, fiber yarn laying pattern, extrusion width, layer height, and flow percentage, were investigated and their optimal values were reported. The attained experimental results showed that, for polylactic acid-glass fiber yarn reinforced composite, with the extrusion width of 0.3 mm, the layer heights of 0.22 mm, flow percentage of 0.43, and the rectangular laying pattern, approximately 50% fiber-volume content can be achieved which resulted in tensile yield strength and modulus of 478 MPa and 29.4 GPa, respectively. There was an excellent agreement between these experimental results and predicted theoretically values by rule of mixture.