Tuning the infill pattern is one of the key features in additive manufacturing to optimise part weight. In this work, the effect of the infill strategy, including rate and pattern type, is studied on the mechanical performance of polylactic acid (PLA)-carbon composite. In particular, three types of patterns and four filling levels are combined. These combinations are evaluated by tensile loading applied on dogbone specimens. In addition, the underlined deformation mechanisms are further explored using filament-based finite element model. The numerical simulation is built from sliced models and converted into 3D meshes to predict tensile performance. The results show that the infill rate has a nonlinear effect on the density of PLA–carbon composites, and its magnitude depends on the complexity of the generated pattern. In addition, tensile loading is found to activate varied modes of shearing and uniaxial deformations depending on the pattern type. This leads to different profiles and rankings of the tensile performance and allows the infill strategy to significantly affect the part performance, along with its density.