In a series of four behavioural experiments, we systematically varied the type of material (meaningful object images vs non-meaningful “squiggle” shapes), the format of the test used to probe memory (a more visually driven alternative forced choice test vs a remember-know recognition test), and the delay of the final test (immediate vs 1 week delay). We found that abstract squiggle images never showed a significant testing benefit, irrespective of test format, and even benefitted more from restudy than retrieval at longer delays, where testing effects are typically most prominent. Meaningful object images did benefit from testing, particularly at long delays, and with a test format probing the recollective component of recognition memory. Together, our results indicate that retrieval enhances memory for visual materials only when they have a unique, distinctive meaning. These results agree with cognitive and neurobiologically motivated theories proposing that retrieval’s benefits emerge through spreading activation in pre-existing semantic networks, producing better integrated and more easily accessible memory traces.