In this study, our group aimed to determine whether there is a relationship between hair color and perceived attractiveness. Although other studies had suggested juxtaposing results, we hypothesized that blonde women would be rated most attractive due to the character traits that blondes are often associated with. Accordingly, we tested 20 voluntary female participants to rate 35 female faces with blonde, brunette, and red hair colors using CogDemos, Microsoft PowerPoint, and a simple Likert scale method to record the participants’ ratings. The results yielded identical medians, ranges, and interquartile ranges for blondes and brunettes and, although more participants rated brunettes as slightly more attractive than blondes, the results were insignificant. Thus, our experiment revealed no significant relationship between hair color and perceived attractiveness.



          Although there have been many studies conducted pertaining to the relationship between certain attributes that people have and their perceived attractiveness in the opinion of others, ranging from intellectual traits such as intelligence to physical characteristics like weight and height, the relationship between hair color and the perceived attractiveness of a woman differs in relation to other attribute-attractiveness relationships in that there have been multiple studies conducted that suggest different theories regarding which hair color is viewed most desirably. For whereas it is widely accepted that women find men between 5’9” and 5’11” (average to slightly taller than average) to be most attractive (Graziano, 1978) or that men find women who fall between the normal BMI rates of 18 and 25 more attractive than the more slim or overweight physiques (Swami, 2005), studies attempting to discover the most attractive hair color have proven to be less definitive.

For instance, studies conducted in Poland (Sorokowski, 2008) and Great Britain (Swami & Barrett, 2011) suggested findings favoring the blonde hair color is considered to be the most attractive for a woman due to the fact that the blonde hair color is considered to personify youth and openness—traits that have also yielded empirical results that suggest blonde women are approached most often in nightclubs and are tipped most generously at restaurants. However, in contrast to results suggesting that blondes are most attractive, the same Swami & Barrett study exhibits that, while blondes are approached most often in nightclubs, men perceive brunettes to be more attractive and intelligent in comparison to blondes who are perceived as needy and more open to casually spending the night with a man, and an additional study conducted in the Greater London area revealed that women with brunette hair are considered most attractive by European Caucasian men (Furnham, 2008). As a result of the mixed, inconclusive findings in the past regarding women’s hair color and attractiveness, the controlled, generalizable, and straightforward study we have conducted has the potential to clarify—or at least help pave a pathway for—a topic that has simply not been as adequately diagnosed and defined as other attribute-attractiveness relationships.

            Due to the fact that university students often seek youthful, fun-loving companions and partners coupled with past findings that struggle to reveal a consistent relationship between a woman’s hair color and perceived attractiveness, the following hypotheses were formulated:

  • ·Alternative Hypothesis: Participants will rate the pictures of women with blonde hair significantly higher than those with brunette.

Null Hypothesis: The pictures of blonde women will receive the same rating as those of brunettes.

  • ·Alternative Hypothesis: Participants will rate the pictures of women with blonde hair significantly higher than those with brunette.

Null Hypothesis: The pictures of blonde women will receive the same rating as those of brunettes.

  • ·Alternative Hypothesis: Participants will rate the pictures of women with blonde hair significantly higher than those with brunette.

Null Hypothesis: The pictures of blonde women will receive the same rating as those of brunettes.

Alternative Hypothesis: Participants will rate the pictures of blonde women higher than those with brunette.

Null Hypothesis: The pictures of blonde women will receive the same rating as brunettes.




We tested 20 first-year female psychology students from the University of St Andrews who ranged from ages 17-21 with the mean age being 18. Participants volunteered to partake in this study.


The stimuli were taken from the CogDemos computer program, and were then presented in a slideshow format on Microsoft PowerPoint where participants proceeded to rate the stimuli.


            In the study, the images (stimuli) were created using FaceGen. We selected 35 female images, 15 blondes, 15 brunettes, and 5 redheads, for our participants to rate based upon perceived attractiveness. We also used a Likert scale ranging from 1 to 5 with 1 signifying very unattractive and 5 signifying very attractive.  


The independent variable in this experiment was hair color as hair color was manipulated, and the dependent variable is the perceived attractiveness of the images measured by the responses from the participants. Control variables include the gender of our participants (female) and the gender of the images (female). As there was no control group in this experiment and this experiment did not test participants multiple times in a ‘before and after’ scenario for example—all 20 participants rated the images in the same manner as each other—this experiment constitutes as a basic between subjects design. However, a form of counterbalancing was used in that the stimuli were presented in a different order each time in the PowerPoint—hence controlling for the potential of a participant ever consciously or subconsciously discovering a pattern that could skewer the legitimacy of her rating.


            In the experiment, participants—after reading the debriefing form and giving consent to partake in the ethically-approved, confidential study with the option of withdrawing at any time during and after the experiment for any reason—proceeded to sit down before the computer to begin the experiment. The stimuli were individually presented in random order in 5-second intervals—making the experiment with 35 images last just under 3 minutes. The participant would rate each image (based upon the tenets of the Likert scale that were explained on paper and orally in the debriefing session) verbally immediately after viewing the image and the experiment conductor(s) would subsequently record the participant’s response. This process occurred every 5 seconds 35 straight times until the completion of the experiment where the participant was thanked for participating in the experiment and reminded that they had the ability to ask any questions, vocalize any concerns, or still withdraw from the experiment. 



          After conducting the experiment in its entirety, the results revealed that the median rating for all blondes and brunettes was 2, the range for all blondes and brunettes was 4, and the interquartile range for all blondes and brunettes was 2.

            The median rating of blondes and brunettes did however vary from participant to participant. Though many participants had the same median rating for blondes and brunettes, multiple participants had a higher median rating for brunettes than blondes, while one participant had a higher median rating for blondes than brunettes.


Perceived Attractiveness of Blondes and Brunettes Comparison      

After conducting a Wilcoxon T-Test, the result between the comparison of the attractiveness of blondes and brunettes was not significant [Given that the experiment was one-tailed: 5% and T(6) must equal less than 2 in order to be significant, T(6) = 4.5, n.s. is the result]. Thus, based on our study, there is no correlation between a woman’s hair color and attractiveness. (See Appendix A)



          With the aim of intending to clarify whether there is a distinct relationship between a woman’s hair color and attractiveness, in regards to a topic that has produced inconclusive and differing results, our conclusion suggests that there is no distinct generalizable relationship between a woman’s hair color and attractiveness. And, considering how other studies suggest many different conclusions—ranging from blondes or brunettes are deemed most attractive to how people are attracted to partners who have very similar or very different characteristics (Rath, 2006)—the insignificant results suggesting that there is no preference in a woman’s hair color, although not in accordance with our alternative hypothesis, is not very surprising. Ultimately, the results suggest that attractiveness is a culmination of other physical and emotional characteristics rather than solely determined by hair color; so in this sense, the results would be in line with the preferable notion that perceived attractiveness goes deeper than hair color.

            Limitations in our study included a predominantly female participant pool which resulted in having to resort to testing solely females in order to control for gender, stimuli that were significantly less attractive than actual people are, and a participant pool that are relatively homogenous in the sense that they all attend the University of St Andrews and are regarded as young adults. Ideally, we would have selected male participants to rate realistic female stimuli—and the participants would have reflected a wider range of ages and hailed from less similar socioeconomic backgrounds. Although it is difficult to hypothesize how our data was affected by having solely female participants from similar backgrounds, perhaps females are less critical of other females’ hair color than men are or young adults are generally more critical than children or older adults in perceiving attraction for examples, having unattractive stimuli definitely contributed to relatively low ratings from all participants.

While others could certainly replicate our study while dually eradicating our methodological shortcomings, the relationship between hair color and attractiveness appears to be rather subjective. Perhaps further research could test how hair color is perceived by certain demographics and ethnicities—potentially revealing that the relationship is quite culturally specific. Additionally, further research could also see if a certain hair color is preferred in accordance with another intangible, self-described character trait—such as brunette and spontaneous or blonde and confident for examples—in order to further try and discover what the ideal woman is like for the largest number of people. By adding more elements to the traditional inconclusive hair color study, further research actually has the potential to make some relatively groundbreaking, conclusive discoveries.  



Furnham, Adrian. (2008). The Influence of Skin Tone, Hair Length and Hair Color on Ratings of Women’s Physical Attractiveness, Health and Fertility. Scandinavian Journal of Psychology, 49(5), 429-437.

Graziano, W. G.. (1978). Height and attraction: Do men and women see eye-to-eye?, Journal of Personality46(1), 128-145.

Rath, Jaclyn. (2006). The Effect of Similar Physical Features on Perceived Levels of Attraction, Community of Undergraduate Journals, 1, 1-9.

Sorokowski, Piotr. (2008). Attractiveness of Blonde Women in Evolutionary Perspective: Studies with Two Polish Samples, Perceptual and Motor Skills, 106(3), 737-744.

Swami, Viren. (2005). Male physical attractiveness in Britain and Malaysia: A cross-cultural study, Elsevier Body Image, 2, 383-393.

Swami, Viren & Barrett, Seishin. (2011). British men’s hair color preferences: An assessment of courtship solicitation and stimulus ratings, Scandinavian Journal of Psychology, 52(1), 595–600.