<p>Unmanned aerial vehicle (UAV) swarms are experiencing a remarkable growth as powerful assets in search and rescue (SAR) missions, offering rapid situational awareness and wide-area coverage in environments that are often inaccessible or unsafe for human responders. Localization, the ability of each UAV to accurately determine its position, and the relative positions of the other members under extreme and uncertain conditions, stands at the core of swarms’ efficiency. This paper presents a detailed review on localization and sensing technologies designed or adjusted to support UAV swarms in SAR environments. We follow the literature fundamental categorization into absolute, relative and cooperative localization frameworks, covering methods that rely on visual, radio and radar sensing, hinting on emerging learning-based multimodal fusion strategies. Particular emphasis is placed on how these methods address the distinct operational constraints of SAR missions, including (i) unstructured and highly dynamic environments such as rubble or flooded terrain, (ii) time-critical limitations, (iii) GNSS-denied or degraded conditions, (iv) intermittent communication links, and (v) critical mission demands. By analyzing localization accuracy and computational trade-offs under such environments, the review highlights integration challenges between sensing, communication and coordination layers. Finally, we identify key research gaps-notably the need for lightweight, cooperative localization based on confidence and robust multimodal fusion for degraded visual scenes, which have to be addressed to enable resilient UAV swarm deployment in real-world SAR operations.</p>