Permanent activation models of conflict tasks assume that the distractor constantly provides input into the response selection process. Consequently, the compatibility effects observed in these tasks should increase at longer RTs, reflected in positive-going delta functions (pDPs), especially for longer stimulus durations However, in a classical Simon task, negative-going delta plot slopes are typically observed, and further, such slopes are unaffected by stimulus duration. This observation is consistent with transient activation models such as the diffusion model for conflict tasks (DMC, Ulrich et al., 2015), which assume a brief and short-lived distractor activation. Since these models describe general cognitive architectures, it is important to test their predictions across various boundary conditions. To that end, we investigated in how far the assumption of a transient rather than permanent distractor activation generalizes across different conflict tasks. Specifically, we compared the effects of short (200 msec) versus response-terminated stimulus duration on the DP slopes in a classic Simon and accessory Simon task (Exp. 1), and in a classic Simon and Eriksen flanker task (Exp. 2). Both experiments revealed that the DP slopes were largely unaffected by stimulus duration. Specifically, the typical nDPs for the classic Simon but pDPs for the Eriksen and accessory Simon task were observed in both stimulus duration conditions. Additional modeling revealed that the parameter of DMC reflecting the time-course of distractor activation varied in accordance with the empirical DP slope pattern. Taken together, we argue that our results suggest a transient rather than permanent distractor activation in conflict tasks.