This study aimed to manufacture PAN-based conductive yarn using a wet-spinning process. Two types of carbon nanomaterials, multiwall carbon nanotubes (MWCNT) and carbon nanofiber (CNF), were used alone or in a mixture. First, to derive the optimal composite solution condition for the wet spinning process, a composite solution was prepared with carbon nanomaterials of the same total mass weight (%) and three types of mechanical stirring were performed: mechanical stirring, ultra-sonication, and ball milling. A ball milling process was finally selected by analyzing the viscosity. Based on the above results, 8, 16, 24, and 32 wt% carbon nanomaterial/PAN composite solutions were prepared to produce wet spinning-based composite films before preparing a conductive yarn, and their physical and electrical properties were examined. By measuring the viscosity of the composite solution and the surface resistance of the composite film according to the type and content of carbon nanomaterials, a suitable range of viscosity was found from 103 cP to 105 cP, and the electrical percolation threshold was from 16 wt% carbon nanomaterial/PAN, which showed a surface resistance of 106 Ω/sq or less. Wet spinning was possible with a PAN-based composite solution with a high content of carbon nanomaterials. The crystallinity, crystal orientation, tenacity, and thermal properties were improved when CNF was added up to 24 wt%. On the other hand, the properties deteriorated when CNTs were added alone due to aggregation. Mixing CNT and CNF resulted in poorer properties than with CNF alone, but superior properties to CNT alone. In particular, the electrical properties after incorporating 8 wt% CNT/16 wt% CNF into the PAN, 106 Ω/cm was similar to the PAN-based conductive yarn containing 32 wt% CNF. Therefore, this yarn is expected to be applicable to various smart textiles and wearable devices because of its improved physical properties such as strength and conductivity.