<p>Cell fate decision making is a core issue in systems biology with profound implications for cellular development and disease. Although dynamical system approaches using gene network models have advanced our knowledge of cell fate transitions, accurately and stably controlling these transitions remains a great challenge. Here, we present a landscape control (LC) approach based on energy landscape theory, which manipulates specific gene targets to direct cell fate. Through testing on a two-gene mutual inhibition and self-activation (MISA) model, an epithelial-mesenchymal transition (EMT) network, and a human embryonic stem cell (HESC) network, we demonstrate that LC significantly outperforms the previous optimal least action control (OLAC) approach in both effectiveness and computational efficiency. Moreover, LC can identify key transcription factors and integrate sparse control strategies to induce specific transitions. Overall, the LC framework provides a valuable tool for studying and engineering cell fate, with potential applications in therapeutic innovation and regenerative medicine.</p>