A co-channel talker interference suppression system processes an input waveform containing the sum of two simultaneous speech signals, referred to as the target and the jammer, to produce a waveform estimate of the target speech signal alone. This paper describes the evaluation of a simulated suppression system performing ideal suppression of a jammer signal given the voicing states (voiced, unvoiced, silent) of the target and jammer speech as a function of time and given the isolated target and jammer speech waveforms. By applying suppression to select regions of jammer speech as a function of the voicing states of the target and jammer, and by measuring the intelligibility of the resulting jammer suppressed co-channel speech, it is possible to identify those regions of co-channel speech on which interference suppression most improves intelligibility. Such results can help focus algorithm development efforts.