BACKGROUND: Arm and hand gesture has been considered a potential facilitator of word production (Skelly, Schinsky, Smith, and Fust, 1974), and gesture is often considered as a therapeutic modality for the treatment of aphasia (Rao, 1994), but there is limited empirical evidence of the efficacy of gesture-based treatments. Models of the relationship between word production and gesture production have been developed (Hadar and Butterworth, 1997; Krauss & Hadar, 1999) but they are currently under-specified and provide little guidance as to whether gesture might be an efficacious treatment for word production deficits arising from particular underlying levels of impairment. AIMS: This study had two main aims: First, to examine the comparative facilitation effects of gesture production and visualisation processes on object naming skills, and second, to compare the effectiveness of three types of treatment, gesture, verbal, and combined verbal plus gesture, for word production deficits arising from impairment at the level of phonological access and encoding. METHODS AND PROCEDURES: A 68-year-old female, AB, participated in the study. AB sustained a single, left, frontoparietal, subarachnoid haemorrhage 6 months prior to the study, which resulted in a highly specific, mild, phonologic access and encoding impairment. AB initially participated in a trial comparing the relative effectiveness of gesture and visualisation processes for facilitating oral picture naming. A controlled multiple-baseline single-case experiment was then carried out comparing the three naming treatments. OUTCOMES AND RESULTS: The use of iconic gesture was found to significantly facilitate picture naming. Pointing, visualisation, and cued articulation produced negligible change from baseline rates. Clinically and statistically significant treatment effects were found for all three treatment conditions, with only marginal differences between conditions. Improvements made in picture naming were maintained at 1 and 3 month follow-up assessments and generalisation of enhanced object naming was found with novel stimuli and during spontaneous conversation. CONCLUSIONS: The results supported Krauss and Hadar's (1999) model of speech and gesture production, suggesting frank interaction between the kinesic monitor of the gesture production system and the formulator of the word production system. The results caution clinicians to question the long-held axiom of the superiority of multi-modality treatments, and encourage clinicians to consider the underlying knowledge and processes generated by particular treatment protocols, rather than simply the modality in which the treatment is transmitted.