Abstract: Monitoring is an absolutely necessary means to maintain the execution reliability of business process in the research on inter-organizational workflow management. Though some work has been done in this field, neither the conventional methods nor the current study on fault tolerance and exception handling in mobile-agent-based systems can effectively solve the problem in the multi-agent cooperatively concurrent process. The migrating workflow system just represents that kind of process. A migrating workflow is performed by several migrating instances, which generally are mobile and possess hierarchical relationship. Hence it becomes difficult to monitor a business flow and recover it from exceptions or faults. To solve the problem of process monitoring and cooperative recovery, a novel hierarchical collaborative monitoring model is presented based on the migrating workflow framework. The migrating instance and some relations between two migrating instances are described. Some constraints are discussed when recovered from faults. The mechanism can supervise migrating instances of different hierarchies by dispatching monitors with corresponding levels. Moreover, the cooperation mechanism among monitors can confine the incidence when facing exceptions. As shown in the analysis and experiment results, the model can implement concurrent monitoring and hierarchical exception dealing. Additionally, it can avoid the inefficiency of single point bottleneck and enhance the reliability of workflow execution.