In two prior studies, we observed that the subjective liking acquired through evaluative conditioning affects visual search times, in particular when set size is changing (presented at TeaP 2022, Cologne). In the present study, we test whether conditioned valence affects visual feature and conjunction search to the same extent. Specifically, if evaluative conditioning induces a bias in bottom-up attentional control, (i.e., changing the stimulus salience), then it may be expected to affect conjunction search more than feature search because feature search is already characterized by clear salience difference between the targets and distractors. Therefore, stimuli were paired with either positive, negative or neutral images in an evaluative conditioning phase prior to being used as target stimuli. In the subsequent visual search task, sixteen stimuli were placed on the screen containing the target or not. In the feature search condition, the target differed from the distractors in only one feature (e.g. colour), while two features were relevant to identify the target in the conjunction search task. As a result of successful evaluative conditioning, the acquired valence of targets was either positive, negative, or neutral, whereas distractors had not been presented previously to the participants. Positively evaluated targets were found slightly faster than neutral and negative targets in the visual search task, and this effect did not differ as a function of whether it was a feature or conjunction search task. Limitations of the study, such as the conditioning material used, are discussed.