Many physiological events require an accurate timing signal, usually generated by neural networks called central pattern generators (CPGs). On the other hand, properties of neurons and neural networks (e.g. time constants of neurons and weights of network connections) alter with time, resulting in gradual changes in timing of such networks. Recently, a synaptic weight adjustment mechanism has been docu¬mented in some synapses, based on timing of pre- and post-synaptic electrical activities, called s spike-timing-dependent plasticity (STDP). We present a model in which the accuracy of the timer network sig¬nificantly improves by using this mechanism. In our model, based on leaky integrate and fire elements, we used 30 timer neurons (time constant=0.0005, resting voltage=-60 mv and threshold=-50 mv) and one feedback neuron (Time constant=0.001, resting voltage=-60 mv, threshold=-50 mv). Some simulated noise was applied to the synaptic connections (random deviation up to 20% of the default synaptic weight). We applied STDP to feedback neuron-timer neurons connections (for both long term depression and long term potentiation: dw=0.01×e-(ISI/100+1) and examined the inter-spike interval (ISI) of the feedback neuron as our model output. There was a significant reduction in ISI variation with (Mean ISI=230 and SD=2.21) and without (Mean ISI=204, SD=10.1) applying STDP. In this simulation, weak synapses will be strengthened because their post-synaptic timer neuron will fire after feedback neuron (long term potentiation) and strong synapses will be weakened because their post-synaptic timer neuron will fire before feedback neuron (long term depression). Therefore, the effect of noise would be partially compensated.