New research published in the British Journal of Psychology provides evidence that physical attraction is not simply about being objectively good-looking. Instead, attraction seems to arise from a complex mix of shared traits, personal preferences, and subtle non-verbal signals that go far beyond facial features. The study examined how people rate others’ attractiveness based on face, voice, body motion, and scent. While some patterns of agreement emerged, much of what made someone attractive varied across individuals.
Traditionally, studies on attractiveness have focused heavily on faces, often identifying broad preferences for features like facial symmetry, averageness, and sexual dimorphism. These patterns were interpreted as indicators of biological fitness and often presumed to hold across individuals and contexts. But attraction in the real world is rarely limited to a headshot. People communicate with their voices, bodies, and even their body odor, all of which may contribute to how they are perceived.
The researchers aimed to move beyond this narrow focus by examining multiple non-verbal cues simultaneously. They also wanted to understand whether patterns of attractiveness are truly universal or shaped by personal compatibility. In particular, they were interested in whether same-sex judgments—which are typically more platonic—differ from opposite-sex judgments, which tend to be influenced by mate selection.
The researchers recruited 61 young adults to serve as “non-verbal agents”—individuals who provided various sensory stimuli for evaluation—and 71 other young adults to serve as “perceivers” who rated these stimuli. The agent group included both men and women who were photographed, recorded speaking, filmed making neutral expressions, and asked to provide body odor samples by wearing sweat-collecting pads under their arms while exercising. This process was designed to capture information across four sensory channels: sight (photos and video), sound (voice), motion (video), and smell (body odor).
The perceivers each rated eight agents (a balanced mix of male and female) across these modalities. Each modality was presented separately to avoid cross-contamination of impressions. For example, perceivers smelled odor samples in isolation and were later shown silent videos or heard voice recordings. They rated each stimulus on a seven-point attractiveness scale, alongside questions about emotional impressions and other subjective reactions. However, for this study, only the attractiveness ratings were analyzed.
The researchers used advanced statistical modeling to examine three main questions: First, do different modalities vary in how attractive they are perceived? Second, are certain modalities redundant—meaning, does finding someone’s voice attractive predict finding their face attractive? Third, do common or personal preferences dominate attractiveness judgments, and does this vary between same- and opposite-sex interactions?
Across the board, the researchers observed that multimodal stimuli—videos with synchronized audio—were rated as most attractive, suggesting that richer, more dynamic inputs provide more favorable impressions. At the other end of the scale, body odor stimuli were rated least attractive on average. Faces, voices, and silent videos landed in the middle.
However, the variation in attractiveness scores between modalities was relatively small. The biggest difference observed between any two modalities was less than one point on a seven-point scale. This suggests that while people have modest preferences for certain types of input, no single channel dominates the attractiveness landscape.
There was also evidence that some modalities carried redundant information. For example, people who were rated as attractive based on their voice were often also rated as attractive based on their appearance. This pattern held across both opposite-sex and same-sex judgments.
The strongest associations were between face and voice, while the weakest were between body odor and photo. However, body odor attractiveness did show a consistent positive link with how someone moved—implying that more dynamic cues like scent and motion may reflect overlapping traits, such as health or hormonal state.
Interestingly, the study also revealed that the weight of personal preference varied by context. When people rated opposite-sex individuals, the variation in ratings could be explained equally by traits of the person being rated (common preferences) and the rater themselves (personal preferences).
In contrast, when participants rated same-sex individuals, the variation was more strongly shaped by personal preferences. This suggests that in same-sex interactions, which tend to be more platonic, subjective compatibility may play a larger role than universally agreed-upon features.
This finding provides support for the idea that attractiveness serves multiple functions. While traditional theories focus on mate selection, the researchers argue that attraction also influences broader social outcomes, such as friendship formation, team cohesion, or general likability. Previous studies have shown that attractive people are often judged as more trustworthy, intelligent, and socially competent—a phenomenon known as the halo effect. These impressions can shape interactions in many social domains, not just romantic ones.
The research took a comprehensive look at various non-verbal cues, but it is not without its limitations. All participants were young adults in good health, and the study was conducted in a controlled laboratory setting. These constraints limit how generalizable the results are to other age groups, health statuses, or more naturalistic social contexts.
The researchers also focused on isolated impressions rather than interactions, so it remains unclear how these non-verbal cues play out in ongoing relationships. Future research might examine how dynamic impressions change over time or explore how shared attraction influences social bonding in real-world settings.
Another limitation was that the study did not measure how attraction relates to behavioral outcomes. While certain traits were judged as attractive, the researchers did not examine whether these traits led to greater cooperation, friendship, or romantic interest in actual interactions. Addressing this gap could help clarify the functional consequences of perceiver-driven preferences.
The researchers also noted the need for further investigation into the specific components of motion and scent that contribute to attractiveness. For instance, future studies might use point-light displays to isolate movement style or analyze odor samples for chemical composition.
The study, “Attraction in every sense: How looks, voice, movement and scent draw us to future lovers and friends,” was authored by Annett Schirmer, Marcel Franz, Lea Krismann, Vanessa Nöring, Marlen Große, Mehmet Mahmut, and Ilona Croy.