Stroop-like stimuli were presented to either the left or the right visual half-field. Subjects responded to the identity of the words above and below (the target dimension), which appeared above or below a reference point (the cuing dimension). Automatic Stroop-like effects were assessed as the difference in reaction times between congruent trials (e.g., above the reference point) and incongruent trials (e.g., above below the reference point) when both trial types were equally frequent. In blocks in which most trials were of one type (e.g., 80% congruent trials), controlled Stroop-like effects could be assessed. Automatic Stroop-like effects remained unchanged under different task manipulations. In contrast, controlled Stroop-like effects were reduced by lowering cue-response compatibility and by increasing the response alternatives from two to four. Thus, similar to other cuing effects, controlled Stroop-like effects are susceptible to manipulations that affect the response-decision stage and appear to involve response-selection processes. The resources supporting these response-selection decisions were not hemisphere-specific, and were sufficiently nonspecific that interference from a memory-load task was found. When resources were scarce, a consistent bias to attend to stimuli presented or responded to on the right was evident.