The aim of the present study was to clarify the conditions under which anodal tDCS applied to left hemisphere language sites may
facilitate picture naming latencies in healthy young adults. We built upon previous studies by directly testing for item-specific and
generalised effects of tDCS through manipulation of item-familiarisation and through testing for both online and offline effects of
stimulation, in the same paradigm. In addition, we tested for the robustness of these effects by comparing two left hemisphere
sites critical for lexical retrieval. Twenty-eight healthy young adults completed two testing sessions receiving either anodal
(1.5mA, 20mins) or sham stimulation (1.5mA, 30s) in each session. Half of the participants received tDCS over the left inferior
frontal region and the other half over the left posterior superior temporal region. All participants were asked to a name a set of
pictures and their response latencies were compared at three time points (before, during and after the end of stimulation). The
stimulus set was constructed so that some items were presented at all time points, some before and after stimulation, and some
during stimulation only. A parsimonious linear mixed effects model revealed robust repetition priming effects as latencies were
reliably faster for previously named items in all conditions. However, active tDCS did not produce any additional facilitation in
relation to sham, and even led to slower performance in the IFG group when the stimulated items differed from those tested at
baseline and post-test. Our findings add to the present debate about the efficacy of single-session tDCS for modulation of lexical
retrieval in healthy young adults. We conclude that future research should take a more systematic, step-wise approach to the
application of tDCS to the study of language and that more sensitive experimental paradigms, which include a training element,
are more adapted to the study of cognitive processes in populations with optimal levels of cortical excitability.