Thalamocortical problems, which are written as systems of Volterra integro-differential equations, are studied. Waveform relaxation, which can be efficiently implemented in parallel computing environments, is applied to the systems. A new error bound is derived for waveform relaxation applied to the thalamocortical problems. Exact values of the kernel of the Volterra integro-differential equations are used in the derivation. Fast convergence of waveform relaxation applied to the problems is illustrated by numerical experiments.