In a recent PNAS article, Chan and LaPaglia (2013) provided arguments and evidence to support the claim that reactivating a witnessed memory (by taking a test) renders the memory labile and susceptible to impairment by subsequent misinformation. In the current article, we argue that Chan and LaPaglia's (2013) findings are open to alternative interpretations, and further test the hypothesis that reactivation increases a witnessed memory's susceptibility to impairment. To this end, the current studies used a different set of materials and a different measure of memory impairment, the Modified Recognition Test (McCloskey & Zaragoza, 1985). In Experiment 1a, we established that our reactivation manipulation was effective by showing that we could replicate the well-established retrieval enhanced suggestibility effect with our materials. However, when we assessed potential impairment of the witnessed memory with the Modified Recognition Test (Experiments 1a and 1b), we failed to find evidence that reactivating the witnessed memory prior to misinformation impaired memory for the originally witnessed event. In Experiment 2, we replicated Chan and LaPaglia's (2013) findings when we used their memory impairment measure (misinformation-free True/False Recognition Test) and showed why that test does not permit clear inferences about memory impairment. Collectively, the results showed that, although the reactivation manipulation increased susceptibility to suggestion (i.e., as evidenced by increased reporting of suggested misinformation), there was no evidence that reactivation through testing increased the original memory's susceptibility to impairment.