The invention discloses an abnormal state monitoring and self recovery method for a scheduling master station front-end processor. The method comprises the following steps: reading system pre-configuration, a network card state and a pre-state, reading all kinds of monitoring configuration parameters, and reading the time of automatically restarting a pre-application last time; creating a monitoring thread, and judging whether the processor sends real-time data to an SCADA normally; periodically monitoring whether a real-time database can correctly open and read the data; periodically monitoring the operation state of a pre-core process; periodically monitoring the working condition of a channel connected to the processor; periodically monitoring the pre-state of the processor, the network card state of the processor and the states of other front-end processors in the system; and carrying out diagnosis by integrating the monitoring results. By adopting the abnormal state monitoring and self recovery method disclosed by the invention, various links in pre-operation can be automatically judged, the monitored abnormality can be automatically processed and initiatively recovered, the requirements of quickly recovering onsite operation are satisfied, and the data collection reliability and robustness of a scheduling master station automation system are improved.