This study investigates how lighting quality and emerging intelligent technologies influence pedestrian safety at designated crossings, with particular focus on cognitive aspects of perception, decision-making, and situational awareness. The aim is twofold: to evaluate whether existing crossings in Hungary meet established lighting standards and to explore how adaptive, sensor driven systems can enhance cognitive safety. Field measurements were conducted at 14 non-signalized crossings in three Hungarian cities—Veszprém, Budapest (District 22), and Békéscsaba—following the MSZ EN 13201 protocol. Accident data from the national WEB-BAL database guided site selection, emphasizing high-risk nighttime locations. Results reveal that only five sites met both recommended thresholds of 20 lx average illuminance and 0.4 uniformity, with poor waiting-area lighting and asymmetrical designs as recurring deficiencies. The study further highlights the potential of adaptive LED systems and smart crosswalk technologies (e.g., SafeCross, SafeXOne) to reduce cognitive load and improve pedestrian visibility. The findings demonstrate that beyond compliance with lighting standards, cognitively responsive infrastructure—integrating intelligent, adaptive solutions—is essential for creating safer pedestrian environments at night.