Background: One application of a task, such as work-picture matching or repetition, has been demonstrated to affect subsequent picture naming ('facilitation' or 'priming') in both aphasic and non-aphasic subjects. As aphasia assessment frequently involves repeated use of the same stimuli in different tasks, it is suggested that some aphasic individuals may have improved performance in picture naming as a result of assessment. Aims: The aim of this paper is to demonstrate that there can be improvements in picture naming as a result of repeated presentation of stimuli and promote discussion regarding the mechanisms that may have caused such effects and their corresponding implications for treatment. Methods & procedures: This paper describes a single case study of JAW, a man with aphasia. JAW's picture naming had been observed to improve over time while other tasks remained stable. An investigation was performed to identify the source of this improvement. Three treatment tasks were used, attempting to name the picture, reading aloud and delayed copying of the picture names. Outcomes and results: All three tasks significantly improved subsequent picture naming of the treated items despite the fact that no feedback or error correction was provided. It is argued that the source of this improvement is from priming of retrieval of the phonological form. In the 'attempted naming' condition, this priming occurred every time a picture name was successfully produced. As JAW was not perfectly consistent, on each attempt at naming some additional items were primed. Thus, over time an increasing proportion of stimulus items were primed and were hence more likely to be successfully produced. Conclusions: This study demonstrated that for at least one aphasic man, JAW, practice makes (closer to) perfect, even without correction. Not only was there significant improvement from tasks that provided the word form (reading aloud and delayed copying) as has been shown in the past, but also there was the novel finding that simply attempting to name a picture can improve subsequent word retrieval. It is argued that this novel finding could be applicable to other aphasic individuals and has functional significance.