Compensation can be defined as a process by which a system is able to regain its normal state in response to losing a part of its structure. From genes to individuals to communities, compensatory processes ensure functionality, stability and resilience. While there is a rich literature on compensation in natural and semi-natural communities, as well as for crop monocultures, quantification of compensation in the context of diversified agro-ecosystems, such as binary crop mixtures, has remained underdeveloped. We elaborate a method of quantifying compensation which is consistent across systems of only one component (monocultures) and two component-systems (mixtures). Using plant communities and their response to loss of plant density as an example, we apply this concept to a published data set on experimentally grown monocultures and mixtures of beans and oats. Against loss of plant density, oat and bean monocultures compensated 46 % and 33 % of grain yield, respectively. Compensation in the mixture was small (9%) when compensating loss of bean density, but larger (42 %) when bean density was fixed, compensating against loss of oat density. Remarkably, in the case of fixed bean density, the mixtures did not show a higher degree of compensation than the oats monoculture (i.e. where bean density is fixed at zero seeds). Further, we generalize this approach by characterizing four variables, the Contributor (which suffers the loss that needs to be compensated; e.g. plant density), the Reductor (which leads to a reduction in the Contributor; e.g. a mortality factor reducing plant density), the Outcome (which is the variable of interest that directly depends on the Contributor; e.g. grain mass per unit area), and the Compensator (which represents the variable responsible for the compensation, e.g. the grain mass per remaining plant) and summarize this as the ‘CROC’ model of compensation. We provide a framework for quantitative research on compensation in intercropping and other contexts, and its relationship to resilience and stability. The analysis of compensation in mixtures can help to develop the potential of diversification for improving multifunctional performance of agro-ecosystems, especially in response to biotic and abiotic stresses.