The gender balance among the top scorers on Advanced Placement exams tends to track the gender balance among students taking those tests, according to a new study led by the University of Georgia.
That gender gap is currently quite wide.
Across more than two decades of exam data, the researchers found participation in AP tests were often split along gender lines. Girls dominated the languages, humanities and arts, while boys took more STEM (or science, technology, engineering and math) subjects.

In many subjects, the more evenly split the AP test takers were, the narrower the gender gaps among students earning top scores became, the study found.
"Gender stereotypes about ability, like persistent beliefs that boys are naturally better at mathematics, physics, computing or similar STEM courses, or that girls are better at arts, can influence both students' choices on which class to take and how well they do in that class," said Kadir Bahar, corresponding author of the study and an associate professor in the Mary Frances Early College of Education. "Students are more likely to enter and persist in a field when they feel that people like them belong in it."
AP classes are offered in high schools across the country. They are designed by the College Board to match the rigor of college courses and can count toward college credits if students score well on the AP exams at the end of term.
Gender gaps narrowed significantly in some subjects, grew in others
The researchers analyzed more than 70 million AP exam results, covering about 40 million students between 1997 and 2020.
Out of 45 subjects, boys outnumbered girls in many science and math-centered subjects, especially physics and computer science, the study found. For example, the researchers found that for every one girl taking the AP Computer Science A exam, more than four boys were taking the exam.
Girls outnumbered boys in studio art exams by almost three to one, and in English literature exams by almost two to one.
"Think about it: You are going to an AP physics course, and there is only one girl for every three boys. That might give a girl a sense that physics is a boys' room. That subject belongs to them," Bahar said. That feeling may make girls question whether they should take a course, even if it interests them.
The researchers found that male overrepresentation has decreased in many of those STEM-related courses - calculus, computer science and chemistry, for instance - over the years.
You are going to an AP physics course, and there is only one girl for every three boys. That might give a girl a sense that physics is a boys' room.
Kadir Bahar, Mary Frances Early College of Education
AP Calculus AB exam participation, for example, stayed close to evenly split between boys and girls while the gap in Calculus BC narrowed. Statistics also evened out, and German and Latin were roughly balanced throughout.
That was not the case for algebra-based Physics 1, though. It remained male-dominated, and the gap widened. In the years the exam was offered (2015 through 2020), three of every five students who took it were boys.
Studio art continued to skew more female, and that gap also grew. Roughly seven out of every 10 students taking those exams were girls.
"We aren't sure exactly what changed," Bahar said. "But there were nevertheless encouraging signs of change. Gender gaps narrowed significantly in several subjects."
Participation in AP tests and exam scores could be linked
Across many AP subjects, gender representation among exam takers was strongly associated with gender representation among students earning a top score of 5. However, the relationship varied substantially by subject.
In some subjects that maintained a sizable gap in participation, the gender that made up the majority of students also made up the highest scores. But in several subjects where girls were most of the test takers, such as biology, environmental science and world history, boys still outnumbered girls among top scorers.
The gaps remained striking in some subjects. Across the study period, boys were represented among top scorers at roughly four times the rate of girls in Physics 1 and nearly six times the rate in Computer Science A.
"Participation and top achievement are related," Bahar said. "Getting more girls into computer science may help narrow those excellence gaps." Physics was a notable exception, though. There, changes in participation did not track changes among top scorers.
Gendered expectations can accumulate over many years.
Kadir Bahar
Simply increasing participation by one gender won't necessarily close gaps among top scorers, the researchers said.
"Gendered expectations can accumulate over many years," Bahar said. "Parents may perceive sons as more mathematically capable and daughters as more verbally capable. Teachers may unintentionally enforce such expectations, and peer norms can make certain subjects appear more or less socially appropriate for boys or girls."
Making all students feel comfortable to try different subjects is key, the researchers said.
"Equity doesn't mean forcing equal numbers of boys and girls into every subject," Bahar said. "But it means ensuring that gender is not creating unnecessary barriers to opportunity, participation, support or high achievement."
The study was published in the Journal of Educational Psychology and was co-authored by Yasemin Sipahi, a doctoral student in the Department of Educational Psychology, as well as Eter Mjavanadze, Erdogan Kaya and Ummugulsum Aydin.