Recent experimental results show the sensitivity of synaptic plasticity to the timing of synaptic events and postsynaptic firings (spike timing dependent plasticity: STDP). Although a number of studies have been made on STDP, little is known about the effect of STDP on the relation between external input patterns to the recurrent neural network and its output spikes. In this study, we examine this relation by computer simulations of a spiking neural network with STDP. We have found that STDP organizes the neural network to transform an external spatiotemporal input pattern into a synchronous firing, and the synchrony is much dependent on the spatiotemporal structure of the external input. This result suggests that the original of the synchrony propagate through feed-forward network may be generated by the recurrent neural network respond to the external input.