With the rapid growth of urban populations, managing city parking spaces has become a significant challenge. Traditional parking systems are inefficient, increasing traffic congestion, pollution, and wasted time. The Internet of Things (IoT) has emerged as a transformative technology in building smart cities, offering the potential to enhance urban living through automated, data-driven solutions. This paper explores IoT-based smart city parking systems integrated with predictive analytics. We discuss such systems' architecture, functionality, and benefits, including real-time data collection, predictive modeling, and resource optimization. Moreover, we examine case studies of smart city implementations, challenges faced, and future directions for improving smart parking systems through advanced machine learning algorithms, cloud computing, and IoT security enhancements.