Selecting items for designing psycholinguistic experiments can be a very hard and time-consuming process, because of the large number of variables that need to be controlled for. This is clearly the case for picture-naming experiments because, thanks to the collection of psycholinguistic norms on both pictures and their names, a large number of factors that affect naming speed and/or accuracy have been found. In the present study, a Bayesian meta-analysis was performed to determine the extent to which the variables that have generally been considered by researchers as important to control for are indeed worth taking into account. The meta-analysis revealed that most of the variables that are considered in picture-naming studies have a strong or very strong influence on naming speed (image agreement, name agreement, image variability/imageability, age of acquisition, and conceptual familiarity), whereas two variables that are very often taken into account (visual complexity and length) yielded null effects. The results were inconclusive for lexical frequency. At a methodological level, Bayesian meta-analyses constitute a very useful tool for guiding researchers when selecting materials for experiments.