Personality types in dating: what the research says

People put their four MBTI letters in dating profiles and search for which types "match." The test is popular. The research behind it is thinner than its popularity, and the research that psychologists rely on says something less exciting about dating: your personality matters, but it explains a small slice of how a relationship goes.

What MBTI is, and where it stands

The Myers-Briggs Type Indicator sorts people into one of 16 types using four pairs of letters: Extraversion or Introversion, Sensing or Intuition, Thinking or Feeling, Judging or Perceiving. According to Wikipedia's summary, its publisher, The Myers-Briggs Company, calls it the world's most widely used personality assessment (a claim from the publisher).

Here's what researchers have reported:

Types versus scales. In a 1989 paper in the Journal of Personality, Robert McCrae and Paul Costa gave the MBTI and a Big Five measure to 468 adults aged 19 to 93. They found no support for the idea that the MBTI measures two distinct kinds of people on each pair, such as "introverts" and "extraverts" as separate groups. Instead the four scales measure four relatively independent dimensions. Per the Wikipedia summary, most people score near the middle on each scale rather than at the ends, which a type system doesn't predict.
Retest results. The same Wikipedia summary reports that studies found between 39% and 76% of people get a different type when they retake the test after five weeks. The publisher says the instrument meets or exceeds the reliability of other psychological tests, and that it was never meant to predict outcomes such as hiring or life results. The two sides measure different things, so the numbers don't contradict each other.
Dating. A search for this post turned up no large, peer-reviewed study showing that particular type pairings, like "INFJ with ENTP," predict lasting relationships. The one small study found, of 62 couples visiting a clinic (Kong, 2010, in a Korean nursing journal), found no significant differences in marital satisfaction according to whether spouses' types were alike, though it did report some differences in divorce proneness for certain types. It's one small clinic sample, so it proves very little either way.

The MBTI isn't meaningless either. McCrae and Costa found its four scales overlap with four of the Big Five traits. One way to read that is to treat your result as a rough description of preferences, and not a category you belong to.

The Big Five

Most personality researchers use a model called the Big Five. It measures five traits, each on a continuous scale:

Openness: curiosity and appreciation for new ideas and experiences.
Conscientiousness: self-discipline and following through.
Extraversion: energy, sociability and assertiveness. Introversion is the low end.
Agreeableness: kindness, cooperation and trust.
Neuroticism: a tendency toward strong negative emotions such as anxiety or anger. The low end is emotional stability.

Per Wikipedia's overview, the model has been replicated across many cultures, though not everywhere, and most findings rely on people describing themselves, which can be biased. Nobody is "a" type under this model. Everyone has a position on each scale.

What traits predict about relationships

Two large studies speak to this.

In 2010, Dyrenforth, Kashy, Donnellan and Lucas studied married couples in three national samples: Australia (5,278 people), the United Kingdom (6,554) and Germany (11,418). A person's own personality accounted for about 6% of the variation in how satisfied they were with the relationship. The partner's personality accounted for 1% to 3%, mostly through Agreeableness, Conscientiousness and Emotional Stability. And how similar the two partners were explained less than 0.5% once those were counted.

In 2020, Joel, Eastwick, Allison, Arriaga and colleagues pooled 43 longitudinal couples studies from 29 labs and used machine learning to see what predicted relationship quality. The strongest relationship-specific predictors were perceived partner commitment, appreciation, sexual satisfaction, perceived partner satisfaction and conflict. The strongest individual-difference predictors included life satisfaction, negative affect, depression and attachment avoidance and anxiety. Negative affect, a tendency toward feeling low or anxious, is closely related to neuroticism, but the study didn't frame the results in Big Five terms. Relationship-specific variables predicted up to 45% of the variation at the start of studies and up to 18% at the end, and individual differences 21% and 12%. Changes in relationship quality over time were largely unpredictable from the self-report measures.

Limits: both studies looked at people already in relationships, mostly using questionnaires, and both are about averages across many couples. They don't show how personality affects meeting someone or the first few dates, and they can't say how your relationship will go.

So what does that mean for dating?

Personality similarity explained very little. In those married couples, how alike partners were on personality added almost nothing beyond each person's own traits. That's about personality scores, not shared interests or values, and choosing a compatible partner goes into those.
How you treat each other predicted more than traits. In the second study, relationship-specific factors such as perceived commitment, appreciation and conflict predicted more of the variation than individual differences did.
Introversion isn't a flaw or shyness. Extraversion is a trait. Social anxiety and shyness are different things, which social anxiety or shyness explains.
A type is thin evidence about a person. A common-sense rule is to treat a label as a conversation starter, not a filter. Attachment styles covers another popular framework and its limits.

Telling someone "I'm an INTJ" gives them a rough sketch. Telling them what you enjoy and how you handle disagreements usually gives them more to work with.

What to do next

Take a Big Five test once, from a research-based source. Treat the result as a rough snapshot of tendencies, not an identity.
Don't filter by type. If a match lists a type, ask what they like doing on a weekend instead.
Write your three non-negotiables. Examples: wanting kids or not, how you spend money, how you handle conflict.
Ask about conflict on a second or third date. "What happens when you and a friend disagree?" gets a real answer.
Watch behavior over a month. Whether they follow through, show appreciation and respond when you're stressed tells you more than a quiz.
Log what happened. In Run Game, the notes on a session or a conversation are a place to write what actually worked, such as what you talked about and how it felt, instead of a type.

Communication differences in dating is a natural next read.