Treating systemic metastases at the micrometastatic stage is a potential strategy to inhibit cancer metastasis. This study aims to establish an apoptosis sensor‐based platform for rapid, effective, and noninvasive identification of drugs that can inhibit the proliferation of micrometastatic cancer cells. We stably transfected the plasmid DNA encoding the fluorescence resonance energy transfer‐based caspase‐3 sensor into highly metastatic melanoma B16F10 cells. The resulting B16F10‐C3 cells were applied for screening of antiproliferative and proapoptotic drugs in two‐dimensional (2D) monolayer, three‐dimensional (3D) spheroids, and zebrafish xenotransplantation tumors. All studies were conducted in 96‐well plates in a high throughput manner. Fourteen compounds including six chemotherapeutic drugs and eight kinase inhibitors were tested. Thirteen compounds failed the tests due to: Drug resistance, low efficacy, poor pharmacokinetic profile, and/or high side effects to zebrafish. The only compound that passed all tests was pan‐phosphatidylinositol 3‐kinase (PI3K) inhibitor LY294002, which inhibited the proliferation of B16F10‐C3 cells in both 2D and 3D cultures. More important, it significantly reduced the xenograft tumor size in zebrafish by decreasing the viability of metastatic cancer cells. Our study suggests that the PI3K/AKT pathway is a potential therapeutic target for the reactivation of tumor dormancy and proliferation of micrometastases. Moreover, this integrated approach is effective for rapid identification of systemic antimetastases drugs.