Modulation classification plays an important role in various civil and military applications. In this paper, we study the modulation classification for successive relaying multiple-input multiple-output (MIMO) systems with interference. To mitigate the effect of interference from another relay node, we adopt a method that jointly identify the modulation types of the source node and the relay node. Specifically, higher order moments (HOMs) and higher order cumulants (HOCs) of the received signal are used as the discriminating features in random forest (RF) to classify the modulation types. Extensive simulations verify the advantages of the proposed scheme over the traditional scheme that treats the interference as colored noise.