We investigated in what stage (semantic or response) the Stroop effect occurs. Participants had to pronounce some colors of the Stroop stimuli in an unfamiliar way (for example, they had to pronounce “yellow” if the color was red). The method allowed us to create a stimulus that was congruent on the semantic level but conflicted on the response level (the word “red” appeared in red ink, but participants had to say “yellow”). We also created a stimulus that was congruent in the response level but not on the semantic level (the word “yellow” appeared in red ink and participants had to say “yellow”, according to the instruction provided). We realized that the Stroop effect does not occur in the case of semantic congruence even if there is a response conflict. The response conflict, in turn, reduces the Stroop effect but still preserves it on a significant level. Thus, there was no influence on response conflict of the Stroop effect in the case of semantic congruency. The data we have obtained do not match with m