This study aims to predict the strength 3D-printed nanocomposite specimens based on the properties of the nanocomposite filament used in the printer's feedstock. For this purpose, nanocomposite filaments were fabricated and then utilized as input material for the 3D printer to manufacture nanocomposite specimens. The strengths of both produced nanocomposite filaments and printed parts were measured. In the modeling section, an appropriate multiscale model was developed. This model first predicts the strength of the nanocomposite filament and then estimates the strength of the printed nanocomposite specimens. The model covers three scales: micro, meso, and macro. The effect of nanotube length and orientation is considered at the microscale, while their agglomeration is taken into account at the mesoscale. Finally, at the macroscale, the final filament strength is calculated using a genetic algorithm. Moreover, the tensile test results exhibited excellent agreement with the numerical modeling outcomes. Finally, using the filament properties, interlayer adhesion properties, print path adhesion properties, and porosity, the strength of the printed part in longitudinal, transverse, and shear directions was predicted using the finite element model. Comparison of the results with experimental values demonstrated that the model provides acceptable accuracy by less than 4% discrepancy with experimental measurement. Employing a second-degree failure criterion and integrating it with classical laminate theory, the strength of the printed nanocomposite specimens with various printing directions was predicted. The levels of discrepancy between estimated values and experimental observations are between 5% to 10% implying on the very accurate performance of the model.