A metacontrast mask suppresses the visibility of, without influencing the reaction time (RT) to, the target. We investigated whether this dissociation results from a sensori-motor pathway immune to masking effects or from the characteristics of stimulus timing in mutually inhibitory sustained and transient channels. For target visibility, para- and metacontrast yielded the usual U-shaped functions. Peak paracontrast occurred at stimulus onset asynchronies (SOAs) of -150 to -100 ms. RTs were relatively low for metacontrast and did not show a systematic change as a function of SOA. The RT contribution from contour-masking was greatest at an SOA of -150 ms (paracontrast) and declined to near zero in the metacontrast regime. The dissociation between visibility and RT seen in metacontrast did not occur in paracontrast, rejecting the theory that RTs are elicited by a single sensori-motor pathway immune to masking. The dependence of the dissociation on stimulus timing can be explained by RECOD, a dual-pathway model wherein fast and slow activities interact.