A picture naming test (PNT) has long been regarded as an integral part of neuropsychological assessment. In current research and clinical practice, it serves a variety of purposes. PNTs are used to assess the severity of speech impairment in aphasia, monitor possible cognitive decline in aging patients with or without age-related neurodegenerative disorders, track language development in children and map eloquent brain areas to be spared during surgery. In research settings, picture naming tests provide an insight into the process of lexical retrieval in monolingual and bilingual speakers. However, while numerous advances have occurred in linguistics and neuroscience since the classic, most widespread PNTs were developed, few of them have found their way into test design. Consequently, despite the popularity of PNTs in clinical and research practice, their relevance and objectivity remain questionable. The present study provides an overview of literature where relevant criticisms and concerns have been expressed over the recent decades. It aims to determine whether there is a significant gap between conventional test design and the current understanding of the mechanisms underlying lexical retrieval by focusing on the parameters that have been experimentally proven to influence picture naming. We discuss here the implications of these findings for improving and facilitating test design within the picture naming paradigm. Subsequently, we highlight the importance of designing specialized tests with a particular target group in mind, so that test variables could be selected for cerebellar tumor survivors.