The invention discloses a second pulse coding timing method, and relates to time synchronization in a distributed system. The principle is that a main node controls node level, and rising edge interrupt processing is adopted from node to node level when the mode is in a receiving state. The method comprises steps: A, the main node enhances node level at the time t0, and the main node time t0 is broadcasted to all slave nodes in an entire second pulse mode via a communication bus; B, the slave nodes are triggered to be interrupted when the lever is pulled up, and the current slave node time t1 is kept; and c, after the slave nodes receive t0 broadcasted by the main node in the entire second pulse mode, the current slave node time t2 is kept, and a new synchronization time t is obtained after calculation and t=t2+(t0-t1). Compared with the prior art, accuracy timing can be carried out only when the main node transmits entire second pulses to the nodes via the communication bus, other communication hardware does not need to be added, and the cost of time synchronization in the distributed system is greatly reduced.