Multiple description (MD) video coding is one of the techniques used to reduce the detrimental effects caused by transmission over error-prone networks. Several approaches have been proposed for MD coding, where each provides a different tradeoff between compression efficiency and error resilience. This paper presents a hybrid MD coding method. The hybrid MD encoder segments the video along both temporal and spatial dimensions, and generates four descriptions using two motion prediction loops. In case of description loss, the hybrid MD decoder adaptively takes advantages of spatial correlation among residual pixels and temporal correlation between frames for lost data estimation. As a result, better error resilience can be achieved at high compression efficiency. The advantages of the proposed hybrid MD method are demonstrated in the contexts of description loss in ideal channels and in packet loss networks.