Magnetic resonance imaging experiments on synesthesia have provided empirical evidence of structural and functional peculiarities in the synesthetic brain. The majority of this research was conducted on grapheme-color synesthesia (GCS), given that it is one of the most prevalent modalities and due to the fact that visual features are more accessible for scientific studies; on the basis of this knowledge, different explanatory models of synesthesia have been proposed. For example, the Cascaded Cross-Tuning model (CCT) suggests that the neural basis of synesthesia is localized in regions that participate in the processing of the inducer (e.g. a letter) and the concurrent (e.g. a color). The model predicts that these regions are likely adjacent and show increased activation, anatomical variations and/or rapid coactivation in synesthetes (1). Evidence from GCS supports this model, whereas data available from other modalities are not enough to test its predictions. Other models, such as the Conceptual Mediation Model (2) or the Emotional Binding Theory (3,4) predict additional activity in areas that participate in lexical meaning and/or emotion. In order to shed light on this debate, data from other modalities of synesthesia are needed. Interestingly, eleven modalities of synesthesia include odors as inducers or concurrents. Emotions, lexemes, touch, personalities, vision or sound can trigger olfactory synesthesias. Additionally, aromas can trigger flavors, temperature, sounds, touch and visual experiences such as color. So far, olfactory synesthesias have been studied only from phenomenological (5) and behavioral perspectives (6). Thus, in order to provide new empirical evidence about the neural basis of synesthesia, which can help test current explanatory models of the phenomenon, we conducted the first fMRI experiment on olfaction in a group of multiple synesthetes. Twelve participants (6 synesthetes: 3 men and 3 women, mean age 36 ± 22.18; 6 non-synesthetes: 3 men and 3 women, mean age 54 ± 19.28) were presented with two mixed olfactory-trigeminal odors: mint and an aqueous dilution of 1-butanol (4% concentration) through an fMRI compatible olfactometer (7). Each odor was delivered 9 times in two different series and recorded in a single session. An event-related design (stimulus duration: 2 s; stimulus onset asynchrony: 22 s) was used to minimize the effects of habituation. The fMRI data were acquired in a 3.0 T Signa HDxt MR scanner (GE Healthcare, Waukesha, WI, USA) with an eight-channel head coil (GE Coils, Cleveland, OH). Functional images were obtained using a T2* weighted echo-planar imaging (EPI) sequence (echo time = 25 ms; flip angle = 77°; matrix size = 64 × 64; field of view = 22 cm; repetition time = 2 s; number of slices = 23). Twenty-three contiguous ACPC-oriented slices (3 mm thickness) covering the inferior part of the brain were acquired. Fieldmap correction (8,9) was applied in order to minimize the impact of susceptibility artifacts. SPM8 implemented in MATLAB R2014a (Mathworks, Inc.) was used for data analysis. A two-sample t-test analysis (synesthetes vs. controls) was applied in order to observe intergroup differences in olfaction. Several areas of the brain showed intergroup differences (Figure 1: synesthetes > controls; the reverse contrast did not show any significant results). First, areas that participate in high level visual processing, such as the left perirhinal cortex (BA 36) and the right fusiform gyrus (BA 37). Second, several areas that belong to emotional networks, such as the left temporal pole, bilateral Insula (BA 13), the right middle frontal gyrus (BA 47), the left anterior/inferior temporal lobe and the right medial frontal gyrus (BA 10). Third, regions that participate in language processing, such as the right superior temporal gyrus (BA 21) and the left superior temporal gyrus (BA 22). It is interesting to note that these regions exhibit other complementary functions. For example, the left perirhinal cortex participates in memory, the left anterior/inferior temporal lobe provides meaning and the medial frontal gyrus is involved in working memory. The neurofunctional basis of olfactory synesthesias is distributed, including visual, lexical, and emotional areas. These data support the idea that the synesthetic brain is able to activate semantic networks in the absence of purely symbolic stimuli, in consonance with the Conceptual Mediation Model. Additionally, these data reflect the overactivation of emotional networks that may be acting as multisensory integrators between the inducer and the concurrent, in line with the Emotional Binding Theory. These findings suggest that both models may be complementary, and reinforce the idea that meaning and emotion are intrinsically related processes.