Stem cell therapy has emerged as a promising strategy for treating stroke, but it still presents several challenges, such as the optimal administration route or safety concerns. In clinical practice, endovascular delivery of cells is preferred to intracerebral injection. However, circulating stem cells must cross the blood-brain barrier (BBB) and might get trapped in peripheral organs, resulting in an insufficient number of cells reaching the lesion. To optimize their migration across the BBB, we aimed to mimic immune cells as they extravasate into the brain parenchyma following injury. For instance, since the chemokine CXCL12 is released after stroke, we transduced human iPSC-derived neural progenitor cells (NPCs) to overexpress its receptor CXCR4 (present in leukocytes) and demonstrated in vitro their enhanced migration towards CXCL12. Although this is a promising result, it overlooks the complexity of the extravasation process across the BBB. Therefore, we developed a CRISPR activation (CRISPRa) screen to determine which molecules enhance NPC migration toward the stroke lesion. NPCs containing the CRISPRa system were validated in vitro and will be further injected endovascularly in stroked mice for subsequent analysis. Secondly, since stem cells may carry the risk of oncogenic transformation, we introduced two different suicide genes into NPCs to enable the selective removal of dividing and/or all cells by specific substrate administration. The generation and preclinical validation of engineered NPCs with improved brain homing capacity and inducible ablation systems will considerably improve their efficacy and safety, bringing us closer to utilizing NPCs for stroke treatment via endovascular injection.