Pest, disease and natural disasters in oil palm have caused significant income loss to Malaysia. Early detection of pest and disease is one of the best prevention procedures. However, the tools or devices to efficiently detect and monitor the problems are still limited, especially for oil palm plantations. Therefore, it is necessary to develop an airbornebased detection and monitoring study to reduce cost and time, and also to cover a wide-scale oil palm plantation area. This study examines the performance of multispectral image from drones and satellites for aerial detection and monitoring of Ganoderma disease, bagworm infestation and flood hazard in oil palm. The images were used to categorise: a) healthy; b) Ganoderma-infected; c) bagworm-infested oil palm and d) stressed oil palm due to flood area. The difference between categories were assessed using visual interpretation of different band combinations displays. Then comparative statistical analysis was conducted to confirm whether there is a statistically significant difference between the spectral response of each multispectral band. The results suggested that NIR, RE, R and G, R, RE band combinations were significantly distinguished between healthy, Ganoderma-infected and bagworm infested oil palm. Meanwhile, NIR, R and G band combinations provided significant difference between healthy and stressed oil palm in flooded area. Future work will explore deep learning analysis for early detection of pest, disease and environmental hazards in oil palm plantations.