The typical treatment for chronic post-stroke anomia is behavioural speech and language therapy. However, such therapy is not always effective or efficient. Previous research indicates that transcranial Direct Current Stimulation (tDCS) can enhance the effects of behavioural speech and language therapy, yet these findings have been limited by the highly varied protocols used across studies. A comprehensive, longitudinal intervention programme was subsequently devised to investigate the effects of systematically varying the laterality and polarity of stimulation on a range of language measures. Outcomes following active perilesional and contralesional stimulation were directly compared with those obtained following ipsilateral sham stimulation. Chapter 3 demonstrated that combining computer-based repetition therapy with 1mA anodal tDCS delivered to the left frontal lobe of a participant with chronic Broca's aphasia led to significantly greater improvements in treated noun naming accuracy than those achieved following therapy alone. This result is in line with neuroimaging findings linking increased activation in left frontal perilesional regions to post-stroke language recovery. Chapter 4 extended this work by repeating the same tDCS-plus-therapy schedule with a further three participants with differing lesion profiles and aphasia diagnoses. Although significant treatment gains were noted, there were no additional benefits of any form of active stimulation for two of these individuals, and the outcomes for the remaining patient were inconclusive. As such, the results of Chapter 4 clearly highlight between-participant variability in response to tDCS. Chapter 5 documented, for the first time, the extent and nature of response inconsistency in confrontation picture naming across multiple trials. When presented with a large corpus of object images twice, 15 participants named an average of almost 26% of items correctly on one occasion and incorrectly on the other. A wide range of demographic, behavioural and psycholinguistic factors provided an incomplete account of the observed patterns in naming response inconsistency. Finally, Chapter 6 comprised a behavioural case series designed to determine the relative importance of visual speech articulation in computer-based repetition therapy. Five of the six participants responded positively to at least one type of therapy, and all showed the greatest therapeutic gains when therapy included articulatory cues. Links between aphasia classifications and patterns of therapeutic response were complemented by exploratory structural neuroimaging findings indicating that different neural regions may mediate the effects of each type of therapy. The designs of the empirical studies in the current thesis facilitate in-depth analysis of therapy outcomes on a patient-by-patient basis. The findings have considerable clinical applicability, and indicate interesting potential directions for future research.