Background: Word–picture matching tasks have been widely used to assess semantic processing in aphasia, but as yet have received little critical evaluation. Successful performance on a word–picture matching task employs several components of the language‐processing system, including lexical and semantic processing of word stimuli and the visual and semantic processing of picture stimuli. Hence it is not only semantic impairments that can affect performance on this task—breakdown in processing at any point from early auditory or visual processing of the word or visual perception of the pictures can affect accuracy. Consequently, performance on a word–picture matching task might be affected by psycholinguistic variables that pertain to any of these levels of processing, such as imageability, word length, word frequency, and the relationship of the distractors to the target.Aims: This study aimed to investigate the factors affecting word–picture matching performance, using one of the most widely used word–picture matching tasks (Subtest 47, Spoken word–picture matching, from PALPA; Kay, Lesser, & Coltheart, 1992).Methods and Procedures: The performance and error patterns of 54 participants with aphasia and 51 elderly control participants, who had completed spoken word–picture matching (subtest number 47) from PALPA, were evaluated. Correlation and regression analyses were used to investigate effects of psycholinguistic variables on performance (frequency, imageability, number of phonemes, semantic and visual similarity, and word association).Outcomes and Results: No variable was found to significantly affect control performance, due to ceiling effects. Imageability, semantic similarity, and word association affected the aphasic participant group performance. Six of the individuals with aphasia showed a significant effect on performance of at least one of four variables; imageability, semantic similarity, frequency, and word association.Conclusions: This study demonstrates that three psycholinguistic variables significantly affect the performance of both the group with aphasia and some individual participants with aphasia. It suggests that accuracy can be influenced not only by the nature of the relationship of the stimuli within the test but also by the individual level of language processing breakdown. Clinicians and researchers need to be mindful of this when using word–picture matching as the basis of their assessment of semantic processing in aphasia.