Are current fMRI research findings misleading about the nature of human working memory? Though increasingly popular as a technique for studying human working memory, fMRI studies have typically failed to account for any behavioural differences or neuronal activation aroused or suppressed by exposure to the fMRI environment itself. Recent behavioural investigation has found that the noise, posture and space limitations imposed during the fMRI scan distracts attention from the task at hand (Ravicz et al., 2000; Raz et al., 2005; Woollacott & Shumway-Cook, 2002). This study aimed to measure the impact of the fMRI environment upon behavioural performance in the popular verbal working memory n-back recognition task (Owen et al., 2005). Twenty-four normal subjects performed a verbal n-back task in a laboratory and an fMRI simulator. Behavioural performance and reported strategy-use were recorded in both settings. Results indicated that healthy adults showed decrements in response time performance in the scanning environment compared to the laboratory, however, accuracy of response did not differ significantly between settings. All participants reported using a sub-vocal rehearsal strategy in task completion, however, five participants reported using additional strategies (visual mnemonics, vocal rehearsal and chunking). Little variation across setting in strategy-use was found. It is suggested that the fMRI environment is a significant distraction to n-back processing working memory mechanisms, but not a significant disruption to the strategies and mechanisms required for accurate responding.