<p>Language assessment tools have traditionally used black-and-white line drawings. However, recent research highlights color as a key feature to distinguish visually similar concepts. Emojis are simplified and conventionalized visual representations that integrate salient visual-perceptual attributes, such as shape and color, into the representation of concepts. This study aimed to compare semantic and lexical access to living (LTs) and non-living things (NLTs) across black-and-white drawings, colored drawings, and emojis. Ninety healthy adults completed a computerized oral-naming task including 64 stimuli (32 living and 32 non-living items) presented in one of the three visual formats. Accuracy was analyzed using a generalized linear mixed-effects model (GLMM), and reaction times were analyzed using a linear mixed-effects model (LMM). Accuracy analyses revealed a main effect of semantic domain, with higher accuracy for the NLTs domain, and a significant format and domain interaction, but no main effect of format was identified. The disadvantage in naming LTs was most evident for black-and-white drawings and was reduced for colored drawings and emojis. Reaction time analyses showed main effects of domain and format as well as a significant interaction, with slower naming of LTs in black-and-white format. Overall, the results suggest that performance across visual formats was broadly similar when considered globally, but that differences emerged when semantic domain was taken into account. LTs were particularly affected by the absence of color. This study highlights the potential of emojis as an experimental visual format and suggests that their role in oral naming and language assessment should be further examined.</p>