Objective: This paper reviews the effects of coaching on malingering of memory impairment. Method: PsychInfo and PubMed were searched up to March 2015, yielding 27 studies that meet the inclusion criteria. Two meta-analyses were conducted to examine the effects of coaching against naive malingering or best effort control. Results: Effect size was medium for the coached versus naive malingering comparison (Hedges's g = 0.53) and large for the coached versus best effort control comparison (Hedges's g = −1.08). These findings suggest that despite performing significantly better than naive malingerers, coached malingerers can be distinguished from best effort participants. Moderation effects were found in respect of type of measurements (in both comparisons) and type of coaching (in comparison with best effort controls). Warning strategies were found to attenuate symptom exaggeration. Conclusion: The present findings lend further support to the evidence that coaching is generally ineffective in avoiding detection. That said, caution should be exercised to coaching with warning which attenuated the sensitivity of symptom validity measures to memory malingering. These findings suggest that the use of warning may engender a different form of feigning that successfully evades detection.