The Blue Eyes Brown Eyes Exercise: One Classroom, Two Days, Real Prejudice

On April 5, 1968 — the morning after Martin Luther King Jr. was assassinated — a third-grade teacher in Riceville, Iowa named Jane Elliott divided her class of 28 white children into two groups based on eye color and told them that blue-eyed people were inferior to brown-eyed people. Within a day, the superior group was mocking, excluding, and bullying the inferior one. Academic performance changed. Friendships fractured. On day two, Elliott reversed the roles. By the end of the week, her students had learned something about prejudice that no lecture could have conveyed: it feels exactly like this.

The Context That Made the Exercise Possible

Jane Elliott was not a psychologist. She was a third-grade teacher at Community Elementary School in Riceville, Iowa — a small, homogeneously white farming community where her students had almost no direct experience of racial discrimination. After King's assassination on April 4, 1968, Elliott faced a pedagogical problem. How do you teach children what discrimination feels like to a population that has never experienced it and, in Riceville's demographic reality, was unlikely to encounter it through ordinary life?

Lectures about racial injustice were insufficient. Reading about discrimination leaves it safely abstract, a problem belonging to other people in other places. Elliott wanted her students to understand discrimination as an experience, not a concept.

She had been thinking about this problem since 1966, partly inspired by a program on Native American history she had watched on television. She had drafted a rough version of the exercise before King's death. His assassination gave it urgency.

The exercise was not designed with the rigor of a controlled experiment. There was no control group, no randomization, no pre-specified outcome measures. Elliott was a teacher, not a researcher. What she was doing was closer to experiential education or a planned classroom intervention. The distinction matters: the blue eyes/brown eyes exercise is often called an experiment in popular accounts, but its evidential structure is that of a pedagogical demonstration, not a controlled study.

That said, what it demonstrated — vividly, repeatedly, across decades of replications in classrooms and corporate training rooms — constitutes a body of observation that researchers have taken seriously.

Source: PBS Frontline. (1985). A Class Divided [Documentary]. Public Broadcasting Service.

April 5, 1968: Day One

Elliott arrived at school on April 5 and told her third-graders that she had read something important: scientists had discovered that people with brown eyes were more intelligent and better than people with blue eyes. She cited no source. The claim was fabricated entirely.

She then enacted a set of rules corresponding to the hierarchy:

Rule	Applied to	Effect
Sit at the front of the classroom	Brown-eyed children	Spatial reinforcement of status
Get extra time at recess	Brown-eyed children	Tangible privilege
Sit at the back of the classroom	Blue-eyed children	Spatial reinforcement of low status
Wear a cloth collar	Blue-eyed children	Visible stigma marker
Cannot drink from the same fountain	Blue-eyed children	Formal segregation
Cannot play with brown-eyed children at recess	Blue-eyed children	Social exclusion enforced structurally

Elliott reinforced the hierarchy verbally throughout the day, praising the intellectual and behavioral qualities of brown-eyed children and finding fault with the performance of blue-eyed children. When a blue-eyed child made an error, she cited their eye color as the explanation.

The results were rapid and disturbing. Within the first day, brown-eyed children became bossy, arrogant, and willing to enforce the hierarchy. They taunted the blue-eyed children. Blue-eyed children who had been academically strong began making errors. One boy who had never had problems with multiplication tables forgot them entirely under the stress of being labeled inferior.

A boy named Raymond — previously a blue-eyed child who had been good friends with a brown-eyed boy named Russell — got into a fight at recess after Russell taunted him about his eye color. Their friendship, intact that morning, was fractured by afternoon.

Source: Elliott, J. (1977). Eye of the Storm [Classroom footage, ABC News documentary, 1970]. Cited in Bloom, S. G. (2021). Blue Eyes, Brown Eyes: A Cautionary Tale of Race and Brutality. University of California Press.

What Changed Cognitively

The academic performance changes were among the most striking observations. Elliott had timed her students on reading and arithmetic exercises at the start of the year. The brown-eyed students, now labeled superior, completed exercises faster than they had before. The blue-eyed students, labeled inferior, performed worse.

This is a demonstration of what researchers would later call stereotype threat — the phenomenon in which awareness of a negative stereotype about one's group impairs performance on tasks relevant to that stereotype. Claude Steele and Joshua Aronson formalized this concept in 1995 through controlled laboratory research, decades after Elliott's classroom demonstration.

Stereotype threat — the risk of confirming a negative stereotype about one's group, which can produce anxiety, cognitive load, and reduced performance on stereotype-relevant tasks, independent of actual ability.

The mechanism appears to operate through working memory. When a person is aware that their performance might be interpreted as confirming a negative group stereotype, monitoring that awareness consumes cognitive resources that would otherwise support task performance. The result is performance that actually degrades — not because of any real deficit, but because anxiety about appearing to have a deficit is itself impairing.

Elliott's blue-eyed students were experiencing this in real time. They knew they had been labeled as lesser. The label was in the room, enforced by the teacher and their classmates. Under those conditions, performance suffered.

Source: Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of African Americans. Journal of Personality and Social Psychology, 69(5), 797–811.

Day Two: Reversing the Roles

On the second day, Elliott told the class she had been mistaken. Actually, she now said, the research showed that blue-eyed people were superior. The rules reversed. Brown-eyed children donned the collars. Blue-eyed children took the front seats and received the extra recess time.

Several things happened that were qualitatively different from Day One, and the difference is important.

The blue-eyed children, having experienced discrimination the day before, did not become as brutal in their role as superior group as the brown-eyed children had been. They were less enthusiastic about enforcing the hierarchy. Some showed visible discomfort with their newly privileged status. Elliott interpreted this as evidence that the experience of discrimination, even brief and simulated, made children less willing to perpetuate it when they gained power.

The brown-eyed children, now in the inferior role, performed measurably worse than they had in Day One's superior role. The performance degradation matched the pattern of Day One for the blue-eyed group — suggesting the effect was driven by the assigned role rather than by the specific children.

At the end of Day Two, Elliott tore up the collars and told her students the truth: there was no research. The superiority of either eye color was a lie she had told to help them understand what discrimination felt like. The class discussion that followed was, by multiple accounts, one of the most emotionally intense sessions in any third-grade classroom on record.

Source: PBS Frontline. (1985). A Class Divided [Documentary]. Public Broadcasting Service.

The Reunion: Fourteen Years Later

In 1984, PBS's Frontline filmed a reunion of the original 1968 class. The students, now in their late twenties, gathered to watch footage of themselves as eight-year-olds and to discuss what the exercise had meant to them.

The overwhelming majority described it as one of the most formative experiences of their lives. They spoke about it as having given them a visceral, unforgettable frame of reference for understanding discrimination — one that had shaped how they thought about race, hierarchy, and fairness throughout their adult lives. Several reported that they had used the experience as a reference point in parenting their own children.

What was absent was significant distress about having been harmed. The former students did not describe the two-day exercise as psychologically damaging. They described it as educational in a way that reading or lecture had never managed to be.

This is not a controlled outcome study. There is no comparison group of students who received conventional instruction about race and prejudice, and no standardized measures of prejudice across time. The reunion footage is compelling testimony, not evidence.

Source: Peters, W. (1971). A Class Divided. Doubleday. [Later updated as A Class Divided: Then and Now, Yale University Press, 1987.]

What the Research Evidence Actually Shows

Elliott's exercise became widely used in diversity and anti-bias training programs, particularly in corporate settings, through the 1980s and 1990s. The spread created pressure for controlled evaluation of its effects.

The research results are genuinely mixed.

Studies examining the original classroom context — elementary school children — generally found positive effects on racial attitudes. Patricia Devine's research on prejudice reduction suggests that inducing dissonance between explicit egalitarian values and discriminatory behavior can shift attitudes, which is consistent with what Elliott's exercise aimed to do.

Studies examining adult corporate training contexts found more varied outcomes. Some participants became defensive when confronted with the exercise's methods. Others reported that being forcibly placed in a discriminated role felt manipulative rather than educational, and responded with resistance. There is evidence suggesting that poorly facilitated versions of the exercise — particularly those run without Elliott's specific teaching skill — can increase rather than decrease in-group/out-group thinking among adults, because the experience of being the inferior group can activate hostility toward the facilitator rather than empathy toward out-groups.

A 2022 review of diversity training interventions found that experiential exercises like the blue-eyes/brown-eyes exercise had stronger short-term effects on attitude measures than lecture-based approaches but more variable long-term effects, and that facilitation quality was a major moderator.

Source: Lindsey, A., King, E., Membere, A., & Cheung, H. K. (2017). Two types of diversity training that really work. Harvard Business Review.

Source: Bloom, S. G. (2021). Blue Eyes, Brown Eyes: A Cautionary Tale of Race and Brutality. University of California Press.

The Ethical Debate

The ethical questions surrounding the exercise are real, and they have become more prominent as the exercise moved from a one-time pedagogical improvisation by a skilled third-grade teacher into a standardized corporate training product.

The core tension: the exercise works, to the extent that it works, by inducing genuine psychological discomfort. It operates by temporarily making participants feel demeaned, excluded, and powerless. This is not a simulation of those feelings — it is, within the boundaries of the exercise, those feelings. Elliott did not explain her true purpose beforehand because doing so would eliminate the effect. Participants cannot experience the full cognitive and emotional impact of discrimination if they know they are being experimentally discriminated against for educational purposes.

This means the exercise involves deception and the induction of real distress as design features. For adults in corporate training, informed consent is complicated by power dynamics — employees who object to a mandatory training exercise may face professional consequences. For the eight-year-olds in Elliott's original class, parental consent for the specific exercise was not obtained.

Critics, including sociologists Ivor Goodson and Pat Sikes, have argued that the exercise is ethically problematic because participants cannot properly consent to it. Defenders, including Elliott herself, have argued that the discomfort is temporary and that the educational benefit — a genuine experiential understanding of discrimination — justifies it, particularly given the ongoing real-world costs of racial prejudice.

The debate has not been resolved, and the disagreement is legitimate rather than one-sided.

Source: Goodson, I. F., & Sikes, P. (2001). Life History Research in Educational Settings. Open University Press.

Elliott's Subsequent Career and Influence

Jane Elliott left classroom teaching and spent the following decades conducting the exercise in corporate boardrooms, government agencies, and university settings around the world. Her approach became more aggressive with adult audiences — she was less gentle with corporate participants than with eight-year-olds — and earned her both devoted admirers and sharp critics.

She appeared on The Tonight Show with Johnny Carson in 1968, where her demonstration of the exercise drew both national attention and hate mail from residents of Riceville who felt their community was being portrayed negatively. The social cost of the exercise to her professional relationships in Riceville was substantial. She eventually left the community.

In 2020, following the killing of George Floyd, Elliott emerged as a widely shared voice on social media and was cited in discussions about anti-racism education. Her original exercise, more than fifty years old, was again being debated as a model and as a cautionary example simultaneously.

Source: NPR (2020, July 8). We Are Repeating The Discrimination Experiment Every Day, Says Educator Jane Elliott.

Stereotype Threat and the Broader Research Context

The most scientifically durable aspect of Elliott's classroom observations is the academic performance data. The pattern she described — that being told you belong to an inferior group degrades your performance on academic tasks — predated the formal laboratory research on stereotype threat by nearly three decades.

Steele and Aronson's 1995 studies found that Black college students scored lower on verbal tests when the tests were described as ability measures than when they were described as non-evaluative laboratory tasks. The difference was not attributable to actual ability — the groups were matched. It was attributable to the awareness of a stereotype about intellectual performance.

Elliott's third-graders showed exactly this pattern under cruder but in some ways more compelling conditions. The stigma was applied publicly, immediately, and with institutional authority (the teacher). The performance degradation followed within hours. The reversal of roles produced reversal of the performance pattern.

This convergence between an improvisational classroom exercise in 1968 and controlled laboratory research in 1995 is one reason the blue eyes/brown eyes exercise continues to be taught in psychology and education courses despite its methodological limitations.

Source: Steele, C. M. (2010). Whistling Vivaldi: And Other Clues to How Stereotypes Affect Us. W. W. Norton.

Is It Research or Theater? The Scientific Critique of the Exercise

The blue eyes/brown eyes exercise is almost universally described as an "experiment" in news coverage and popular accounts. That framing is inaccurate, and the inaccuracy matters. An experiment, in the scientific sense, requires at minimum a control group, pre-specified outcome measures, and some mechanism for ruling out alternative explanations. Elliott's classroom had none of these. There was no comparison class receiving conventional instruction, no validated pre/post measures of prejudice, and no way to separate the effect of the exercise itself from the effect of the teacher's specific authority, pedagogical skill, or the emotional weight of King's assassination the previous day.

What Elliott ran was a classroom demonstration — closer in structure to theater than to science. The distinction is not a demotion. Theater can teach. Demonstrations can illuminate. But calling this exercise an experiment imports a level of causal certainty the design cannot support. When researchers or trainers cite it as "proof" that discrimination can be manufactured in a day, they are treating observational anecdote as controlled evidence.

Demonstration vs. experiment — a demonstration shows that a phenomenon can occur under certain conditions; an experiment tests whether a specified cause produces a specified effect while controlling for alternative explanations. The blue eyes/brown eyes exercise is the former, not the latter.

The consent and debriefing problems compound this. Elliott did not inform parents in advance of what the exercise would involve. She could not: prior disclosure would have eliminated the effect. But this means the eight-year-olds who experienced real distress, real performance degradation, and real fractured friendships did so without any informed consent framework and with only a brief in-class debrief at the end of Day Two. In adult corporate training settings, the problem intensifies. Employees subject to mandatory diversity training cannot meaningfully refuse participation, which makes "voluntary consent" to the exercise largely fictional.

Critics including sociologist Stephen Bloom, whose 2021 book Blue Eyes, Brown Eyes offered a critical reappraisal of Elliott's career, have argued that the distress the exercise produces is genuine rather than simulated — that participants are not roleplaying discrimination but briefly experiencing it. That is precisely Elliott's pedagogical intent. The ethical question is whether inflicting real distress without full prior consent is justifiable, and whether the person facilitating the distress is qualified to manage the psychological fallout.

What controlled research on prejudice reduction does show is instructive by contrast. Elliot Aronson's Jigsaw Classroom, developed in 1971 to address racial conflict following desegregation in Austin, Texas schools, used structured cooperative interdependence — students in mixed-race groups, each holding unique information the group needed — with measurable outcomes. Studies found decreased prejudice, improved academic performance for minority students, and increased cross-race friendships within six weeks. The effect held across replications. Thomas Pettigrew and Linda Tropp's 2006 meta-analysis of 515 intergroup contact studies confirmed that cooperative, equal-status contact reliably reduces prejudice — a finding that supports the theory Elliott's exercise embodies, through evidence Elliott's exercise itself cannot supply.

The exercise points at something real. The controlled research confirms it. But they are doing different work, and conflating them obscures what each can and cannot tell us.

Source: Bloom, S. G. (2021). Blue Eyes, Brown Eyes: A Cautionary Tale of Race and Brutality. University of California Press — Bloom, a journalism professor, conducted extensive interviews with Elliott's former students and colleagues, offering the most detailed critical account of the exercise's long-term reception.

Source: Pettigrew, T. F., & Tropp, L. R. (2006). A meta-analytic test of intergroup contact theory. Journal of Personality and Social Psychology, 90(5), 751–783 — Meta-analysis of 515 studies covering 250,000+ participants, establishing the evidence base for structured intergroup contact as a prejudice-reduction intervention.

Source: Aronson, E., & Patnoe, S. (1997). The Jigsaw Classroom: Building Cooperation in the Classroom (2nd ed.). Addison Wesley Longman — Describes the original Austin intervention and the subsequent research on jigsaw outcomes across school settings.

Frequently Asked Questions

Q: Was Jane Elliott a psychologist or a researcher?
A: Neither. She was a third-grade elementary school teacher with no formal training in psychology or research methodology. The blue eyes/brown eyes exercise was a pedagogical intervention, not a controlled scientific experiment. Its influence on psychology came through observation and the research it inspired, not through its own empirical design.

Q: Did the children's parents consent to the exercise?
A: Not specifically. Parents were not informed in advance that the exercise would be conducted. Elliott notified the principal after the first day. This is one of the ethical objections to the exercise and distinguishes it from research that would require informed consent today.

Q: How long did the negative effects on performance last?
A: Within the two-day exercise, performance deficits reversed when roles reversed. Elliott's subsequent monitoring of her students did not show lasting academic impairment. The long-term effects appeared to be educational rather than harmful.

Q: Has the exercise ever been replicated in a controlled way?
A: Versions have been studied, particularly in diversity training contexts. The results are consistently more positive in children's settings and more variable in adult settings, with facilitation quality emerging as a major factor. No study has replicated the exact 1968 procedure under controlled conditions with random assignment.

Q: What is the strongest criticism of using the exercise in corporate diversity training?
A: That it induces real psychological distress under conditions where participants cannot meaningfully refuse (particularly when training is mandatory), that poorly facilitated versions can increase rather than reduce prejudice by provoking defensive reactions, and that short-term attitude changes measured immediately after the exercise may not persist in behavior.

Q: How does the exercise relate to the contact hypothesis?
A: It differs significantly from simple contact. The contact hypothesis holds that positive intergroup contact under equal-status conditions reduces prejudice. Elliott's exercise used unequal contact — deliberately manufactured inequality — to create experiential understanding. They target the same problem through opposite mechanisms.

Q: Why did Elliott choose eye color specifically?
A: Because eye color is arbitrary, visible, immutable, and within her class carried no prior social meaning. It satisfied the conditions for demonstrating that any physical characteristic, given the right institutional framing, can become the basis for hierarchy. This is the structural point: the content of the stereotype matters less than the institutional power to enforce it.

The Takeaway

What Jane Elliott demonstrated in 1968 is not subtle: given institutional authority, an arbitrary visible characteristic, and structural rules that create advantage for one group, you can manufacture real prejudice in real people in under a day. The cognitive effects — distorted performance evaluation, academic decline among the stigmatized, in-group solidarity against the out-group — emerge automatically, without anyone choosing them. The exercise also showed, in the reversal, that experience is a more powerful teacher than description. Understanding discrimination is different from having felt it, even briefly, even artificially. That gap between knowing and feeling is the problem Elliott's exercise was designed to close.

---

Read next:

• The Robbers Cave Experiment
• The Milgram Obedience Experiment
• The Little Albert Experiment