Preparation over time is a ubiquitous capacity which implies decreasing uncertainty about when critical events will occur. This capacity is usually studied with the variable foreperiod paradigm, which consists in the random variation of the time interval (foreperiod) between a warning stimulus and a target. With this paradigm, response time (RT) effects of the current and preceding foreperiods are usually observed (respectively called "foreperiod effect" and "sequential effects"). Both single-process trace conditioning mechanisms and dual-process accounts have been proposed to explain these behavioral effects. This study aimed at understanding how manipulations of the inter-trial interval (ITI: 1 s vs. 20 s) and the task context (simple vs. choice RT task) affects the two behavioral effects. Results show that, regardless of the type of RT task, attenuated sequential effects were observed with the longer ITI, contrary to predictions derived from the trace conditioning literature. However, the influence that the ITI duration exerted on the FP effect critically depended on the task context, since the FP effect increased as a function of ITI with a choice RT task but decreased with a simple RT task. These findings support a dissociation between foreperiod and sequential effects, consistent with a dual-process account. (Contains 1 table and 1 figure.)