Garner interference (GI) occurs when certain stimulus dimensions, like the length and width of a rectangle, cannot be independently processed. Ganel and Goodale (2003, Nature) reported that GI occurred in speeded-classification and manual size estimation (ManEst), but not in visually-guided grasping. This was taken as further evidence for the Perception-Action Model, which posits that visual information is processed in a fundamentally different way for actions (like grasping) than for perception (like speeded-classification and ManEst); with vision-for-action being attributed to the dorsal cortical stream and vision-for-perception to the ventral stream. However, in our replications and a meta-analysis, we found that it is surprisingly difficult to replicate GI in ManEst. One possibility to resolve this issue is manipulation of online visual feedback: GI in ManEst was typically tested closed-loop (participants had full vision of hand and stimulus). In the current experiment (planned N=24), we will use open-loop conditions (vision is suppressed after start of the movement). There is some evidence that open-loop conditions favor the occurrence of GI by altering the temporal profile of the responses (Hesse & Schenk, 2013, Behavioral Brain Research). Further, we will include a ‘correlated’ condition to analyze Garner facilitation in addition to GI. Unlike GI, Garner facilitation allows a comparison of conditions with an equal number of stimuli presented per block (Garner, 1970, American Psychologist) and has not been investigated in ManEst. Taken together, we will investigate when and under which conditions Garner effects occur in ManEst, trying to clarify which processing guides vision-for-perception.