Magnetic-activated cell sorting (MACS) is a widely used strategy for isolating defined cell populations from complex biological samples. Conventional MACS primarily distinguishes cells according to whether they are magnetically labeled, which can limit its ability to differentiate multiple populations within a single separation process. To address this, we introduce a separation strategy for broadening the range of transport responses achievable within a magnetic cell separation system. The underlying dynamics are examined through analytical, modeling, finite-element analysis, and time-domain numerical simulations. An experimental platform is developed to evaluate the effects of system configuration, operating conditions, and target properties on magnetically driven transport. The predicted behaviors are experimentally characterized by using controlled synthetic particle systems and biological cells. The results show that the target motion can be effectively modified and tuned using the proposed strategy. These findings provide a general framework for extending magnetic separation beyond its conventional operation and support the development of more versatile approaches to particle manipulation and multi-target separation.