
For about four months I've been quietly tracking something that started as a joke at a dinner. Six of us at the table, and within twenty minutes someone had pulled up a compatibility chart, pointed at me, and said, "Maren, you and Jess shouldn't be close friends on paper." Jess and I had been close friends for nine years at that point.
That small friction got me thinking. I'm an INFJ — supposedly the "rare" introvert everyone writes long Reddit posts about — and most of my closest friendships are with people whose types, according to the charts, should barely tolerate me. So I started paying attention. Not to predictions. To what actually held up.
Here's what I've noticed. And what I no longer believe about MBTI friendship, even as someone who finds the framework useful.

The Myers-Briggs Type Indicator sorts people across four preferences: how you get energy (I/E), how you take in information (S/N), how you decide (T/F), and how you structure time (J/P). Sixteen types fall out of that.
Here's the part most compatibility posts skip: MBTI describes preferences, not behaviors in a specific friendship. My being an INFJ tells you something about how I recharge. It tells you almost nothing about whether I'll text you back on a Thursday.
Where type does show up in friendship, in my experience, is at three specific friction points: energy pacing (how often you hang out and how long), processing style (do you think by talking or by disappearing for two days), and conflict response (direct confrontation vs. avoidance). Two people can share zero letters and navigate these three things cleanly. Two people can share all four and still miss each other completely.
A 2009 peer-reviewed study on extraverted and introverted friendship pairs found that mixed-temperament dyads actually developed richer coping strategies under social pressure than matched pairs. The "incompatible" friendships did more interpersonal work — and it paid off.

I want to be careful here. The pairings below are patterns I've observed and that show up repeatedly in MBTI friendship compatibility analyses — not guarantees. Read them as starting hypotheses about where friction might be low.
INFJ–ENFP. The one everyone writes about, and honestly, it earns it. The ENFP brings scattered curiosity; the INFJ brings pattern-finding. They loop between "what if" and "why does it matter" without getting stuck. My longest-running close friendship is this combination. Week three of knowing each other, we had a four-hour conversation that went nowhere useful. That's the tell.
INTJ–INFP. Both live largely in their heads and don't require small talk to feel close. They can sit in a room reading and call it hanging out. This is the friendship that survives long silences.
INFJ–INTP. Abstract, weird, occasionally awkward. Nobody is trying to fill space. According to research summarized by Psychology Today, introvert-introvert pairs often report feeling less "maintenance pressure" — which matches what I've seen.
ENFP–ENTP. Rapid-fire, argumentative, never bored. These two will plan a trip at 11pm and actually take it. They can also exhaust each other, which is why this friendship often runs in bursts.
ESFP–ESTP. Present-tense people. They don't process feelings by analyzing them; they process by doing something. If your friendship runs on shared activity more than shared introspection, this is the register.
ENFJ–ENFP. Warm, expressive, high-contact. This pair is great at group friendships — they bring others in rather than cocooning.

ISTJ–ISFJ. Quiet, dependable, shows up. They don't need weekly check-ins. They need to know you'll be there in three years. MBTI distribution data suggests these are two of the most common types in the U.S. — so statistically, many long-running friendships are probably this shape.
ESTJ–ISTJ. The "we've been friends since college and still see each other every other month" pair. Low drama, high reliability.
ISFJ–INFJ. Emotionally attentive, both. This one tends to run deep but narrow — two people who know each other's actual life, not just the highlight reel.
This is where I break from most articles I've read.
No MBTI combination is automatically bad. The "challenging" pairs are the ones where the friction points I named earlier — pacing, processing, conflict — land on opposite sides.
Some patterns I've noticed:
None of these are doomed. They require one specific skill: naming the friction out loud instead of letting it accumulate. The friendships I've seen collapse weren't the "incompatible" ones on paper. They were the ones where no one said, "hey, I think we want different things from this."

I'll keep this short because most of it is common sense wearing MBTI clothing:
Worth trying if your setup looks anything like mine: pick one friend whose type is furthest from yours and name one pattern you've noticed between you. Not a diagnosis — a noticing. See what happens.
There isn't a universal answer, and anyone selling you one is oversimplifying. Based on analyses from the 16 MBTI personality type descriptions and my own running list, the pairs that seem to start with the least friction are NF–NF combinations for deep friendships, SP–SP for activity-based ones, and SJ–SJ for long-haul loyalty. But the best friend I have is an ESTP, and I'm an INFJ. Mileage genuinely varies.
Yes, and often they're the most growth-producing friendships. The 2009 study on mixed-extraversion friend pairs found these dyads built more sophisticated communication strategies under stress. Opposite doesn't mean incompatible. It means you'll have to be more explicit about what you need, which — honestly — most friendships should be doing anyway.

Introverted types, particularly INFJs, INTJs, INTPs, and ISFPs, often report smaller friendship circles. This isn't a flaw — research on introversion and friendship maintenance suggests introverts simply trade breadth for depth. Fewer friends, more invested in each.
Not by pretending to be extraverts. The introverts I know who keep long friendships do three things: they communicate that low contact isn't disinterest, they protect one-on-one time over group hangouts, and they're willing to be the one who sends the "hey, still thinking of you" text even if it's been a while. Maintenance for introverts is asynchronous, not absent.
Honestly? It's a starting point, not a verdict. The MBTI has well-documented validity critiques in academic psychology — test-retest reliability is lower than researchers would like, and the binary categories oversimplify traits that are actually continuous. I use it as a conversation opener and a friction-diagnosis tool. Not as a compatibility filter.
Still thinking about why the dinner-table compatibility chart felt so off to me that night. I think it's this: friendship isn't something you match into. It's something you notice held, and then you pay attention to why.
I'll come back to this in a few months once I've tracked more of it.
Previous posts: