The integration of unmanned aerial vehicles (UAVs) into agriculture has emerged as a transformative approach to enhance resource efficiency and enable precision farming. UAVs are used for various agricultural tasks, including monitoring, mapping and spraying of pesticides, providing detailed data that support targeted and sustainable practices. However, effective deployment of UAVs in these applications faces complex control challenges. This paper presents a comprehensive review of UAVs in agricultural applications, highlighting the sophisticated control strategies required to address these challenges. Key obstacles, such as modelling inaccuracies, unstable centre of gravity (COG) due to shifting payloads, fluid sloshing within pesticide tanks and external disturbances like wind, are identified and analysed. The review delves into advanced control methodologies, with particular focus on adaptive algorithms, backstepping control and machine learning-enhanced systems, which collectively enhance UAV stability and responsiveness in dynamic agricultural environments. Through an in-depth examination of flight dynamics, stability control and payload adaptability, this paper highlights how UAVs can achieve precise and reliable operation despite environmental and operational complexities. The insights drawn from this review underscore the importance of integrating adaptive control frameworks and real-time sensor data processing, enabling UAVs to autonomously adjust to changing conditions and ensuring optimal performance in agriculture. Future research directions are proposed, advocating for the development of control systems that enhance UAV resilience, accuracy and sustainability. By addressing these control challenges, UAVs have the potential to significantly advance precision agriculture, offering practical and environmental benefits crucial to sustaining global food production demands.