Long fiber-reinforced thermoplastic (LFRT) composites are a very important lightweight material for practical use by the automotive industry because their mechanical performance is superior to that of short fibers. Microstructures of LFRT products, including fiber orientation state and fiber length distribution, strongly affect their mechanical properties and testing. Prior numerical approaches were limited by their inability to analyze many complex flows. Currently, the 3D numerical technology is able to analyze a varying geometry flow. We therefore aim to derive accurate predictions of fiber microstructure in a reliable 3D flow simulation of an injection molded LFRT composite part, with the use of objective fiber orientation and fiber length models. In addition, the microstructure data imported to calculate appropriate mechanical properties effectively improves the prediction accuracy of structural performance for an LFRT part. In such an integrative simulation, the predictive results are compared to related experimental data for the purpose of evaluation.