16-Type Compatibility Matrix

16 TYPES COMPATIBILITY PERSONALITY RELATIONSHIPS

By RECATOOLS Editorial Checked against sources by Jeffrey Tan, Founder & Architect, RECATOOLS Last reviewed 13 Sep 2026

About this tool

How the 16 types pair up — communication, friction, growth. No fake scores.

How the 16 personality types pair up — communication, friction, what each side brings, and how each pairing grows. Honest editorial profiles, no fake compatibility scores.

RT-PSY-029 · Personality Tests

Pick two personality types and read how the pairing actually works — communication styles side by side, where the friction lives, what each side brings, and how the combination grows. Every page is an honest editorial profile: no percentage scores, no "soulmate" verdicts, and the pairing bands we use to organise the pages are our own disclosed convention, not a measurement.

40 pairings live
More added in curated batches
Communication-first, not horoscope

Don't know your type yet? Take the 16-type test first →

⚠ Disclaimer: FOR ENTERTAINMENT AND SELF-REFLECTION ONLY. NOT A DIAGNOSTIC OR CLINICAL TOOL. This personality assessment uses an original RECATOOLS item set operationalising a public framework — the framework and its originators are cited on this page. Results are educational and reflective in nature and should not be used to make important life decisions about career, relationships, mental health, or hiring without input from qualified professionals. Results reflect self-reported preferences at one point in time and can change on retake, particularly for type-based results near category boundaries. RECATOOLS is not a psychological service provider; no therapist-client relationship is created. If you are experiencing mental health concerns, please consult a licensed mental health professional. Your answers are scored entirely in your browser and are never uploaded or stored by RECATOOLS. Viewing a result page works like any other page on this site and is covered by our Privacy Policy.

The content: Original RECATOOLS editorial interpretation of pairings between the sixteen four-letter Jungian-tradition preference codes. The pairing bands are a disclosed RECATOOLS convention, not a psychometric instrument. Pages are generated from one openly-documented content schema, every page carries hand-written pair-specific analysis, and the build is documented in this tool's provenance record. The sixteen four-letter codes are a popular typology descended from Carl Jung's Psychological Types (1921) through the work of Katharine Briggs and Isabel Myers. MBTI® and Myers-Briggs Type Indicator® are trademarks of The Myers & Briggs Foundation; RECATOOLS is not affiliated with them, these pages are not that instrument, and nothing here is a clinical or diagnostic claim.

📅 Research current as of 13 Sep 2026 · Sources: Typology lineage: Jung, Psychological Types (1921); Myers and McCaulley, MBTI Manual (1985) for the four-letter convention. Reliability limits: Pittenger 2005. The pairing pages are RECATOOLS editorial work — no compatibility scores, no clinical claim, and not the trademarked MBTI® instrument.

Standards, published formulas, and reference data are revised over time — check against a current authoritative source before relying on these figures.

About These Pairings — Method, Honestly Told

How the Pages Are Built

Every pairing page combines three layers. First, the preference math: the two types' four axis letters compared side by side, because the shared letters predict the comfortable parts of a pairing and the differing letters predict both the friction and the usefulness. Second, each type's standing patterns — how it communicates, how it behaves in conflict, what energises and drains it — drawn from the same profiles used across our 16-type test. Third, and most importantly, a hand-written analysis of the specific pairing: the dynamic that emerges when exactly these two patterns share a project, a friendship, or a home. We publish a pairing only when that third layer is written — no page on this site is automated recombination — which is why the matrix grows in batches rather than appearing complete on day one.

The organising bands — mirror pair, kindred pair, balanced pair, stretch pair, full-contrast pair — simply count shared preference letters, four down to zero. They're a disclosed editorial convention for navigation, not a compatibility score, and you should treat any site that gives type pairings a percentage with the suspicion it deserves. Worth knowing about the addresses too: every pairing lives at one canonical page in alphabetical order — INTJ + ENFP and ENFP + INTJ are the same relationship, so they're the same page, and the reversed order simply redirects. Each page covers both directions of the pairing on purpose: what the first type brings the second and what the second brings the first, the friction as each side experiences it, and growth advice addressed to both. Nobody in a pairing is the protagonist here.

The matrix grows deliberately rather than all at once. A pairing page ships only after its hand-written analysis clears our build checks — a minimum depth floor and an originality comparison against both sibling pages and the two parent type profiles — and new batches are added as they're written and reviewed. If the pairing you want isn't live yet, the type index below shows what is, and the fastest route to getting more of your type's pairings published is, honestly, taking the test and sharing it — traffic tells us which corners of the grid to write next.

"No pairing on this grid is doomed and none is guaranteed — the letters predict where the friction will live, not whether two people will do the work. The pages are maps of the terrain, not verdicts about the travellers."

What Pairing Pages Can and Can't Tell You

The honest note, in the same spirit as every page on this site: we know of no peer-reviewed evidence that four-letter type matching predicts relationship success, and the underlying typology has well-documented reliability limits (Pittenger 2005) — while there is plenty of evidence that the things which do predict it (how couples handle conflict, whether partners feel heard and responded to, kindness under stress) are skills and choices available to every combination on this grid. Type vocabulary earns its place differently: it gives two people fast, blame-free language for differences that otherwise get moralised — "you never want to go out" becomes "your battery and mine charge differently", which is a conversation instead of a verdict. Read your pairing for its friction map and its scripts, try the ones that fit, and ignore any sentence — here or anywhere — that tells you who you can or can't love. These pages are written for self-reflection and better conversations, full stop.

Browse the Pairings by Type

Frequently Asked Questions

Deliberately not. We know of no published research that supports scoring four-letter type pairings numerically, so a percentage would be theatre. What the pages give you instead is the useful part: where the friction predictably lives, what each side brings, and the scripts that defuse the classic loops. Any site offering "87% compatible" is making the number up — ours would be made up too, so we don't print one.
The honest answer: research doesn't find one. Relationship outcomes are predicted by how couples handle conflict, repair, and responsiveness — skills available to every pairing — far more than by type matching. "Golden pair" lists are folklore. Use your pairing page to understand a specific relationship you're actually in; don't use the grid as a shopping filter.
Sometimes, and it's neither doom nor magic. Full-contrast pairings (zero shared letters) tend to have more friction surfaces AND more complementary coverage — each partner is strong exactly where the other is blind. Our full-contrast pages treat both sides honestly: the trade is real work for a more complete toolkit, and plenty of couples consider it a bargain.
Because they're the same relationship, so they're the same page. Every pairing lives at one canonical address with the two codes in alphabetical order, and the reversed order redirects there. Each page covers both directions on purpose — what each type brings the other, and the friction as each side experiences it. Nobody in a pairing is the protagonist.
Because we haven't written it properly yet. Every page ships only when its hand-written pair analysis clears our depth and originality checks — no auto-generated filler — so the grid grows in curated batches rather than appearing complete on day one. The index above shows what's live; the busiest corners of the grid get written first.
No, and every page says so. The bands simply count shared preference letters — four down to zero — as a navigation aid. They predict roughly where friction will concentrate (differing letters) and where blind spots double up (shared letters), which is useful framing, but they are our editorial convention, not a psychometric measurement.
All three — and they're written communication-first on purpose. Most of any pairing's dynamics (conflict styles, energy budgets, planning friction) play out identically between colleagues, friends, and partners; the romantic angle gets at most one section per page. Some of the most useful applications are co-founder and teammate pairings.
They get their own pages, and they're the most underestimated on the grid. Communication feels effortless because both run the same operating system — and the real risks live exactly where both share the same blind spots, with nobody wired to cover them. The same-type pages focus on territory, duplication, and deliberately importing what the mirror can't show.
It helps, and it takes three minutes — your code plus honest axis bars showing how solid each letter is. That last part matters here: if you sit near an axis border, two different pairing pages may each be half-yours, and knowing that reads better than treating either as gospel. You can also just look up the types of people you know; the pages work either way.
No — and distrust anything that claims it can. These pages map terrain: where friction tends to live and which scripts help. Whether a specific relationship is right, safe, or worth its work involves the actual people, history, and circumstances — things no type grid can see. For decisions that matter, the pages are conversation fuel, and a licensed counsellor is the professional-grade tool.

Method & sources

How it computes

Editorial pairing pages between the sixteen four-letter preference codes (E/I, S/N, T/F, J/P). Each unordered pair lives at one alphabetical address, and pairs are grouped into five navigation bands by the number of shared letters, four down to zero. No compatibility score is computed and no instrument is administered on this page: the codes are the popular typology descended from Jung's Psychological Types (1921) through Briggs and Myers, not the trademarked MBTI® assessment.

What this tool implements

Curated-only publication: a pairing page ships only with hand-written pair-specific analysis; the band labels are a disclosed editorial convention, not a measurement.
No numeric compatibility claim and no clinical or diagnostic claim; the about section says we know of no peer-reviewed evidence that type matching predicts relationship outcomes and cites the typology's reliability limits (Pittenger 2005).
Trademark protocol: MBTI® and Myers-Briggs Type Indicator® are named once as trademarks of The Myers & Briggs Foundation with a non-affiliation statement.
Both directions of each pairing are covered on one canonical page; the reversed order redirects.

Sources

Jung CG. Psychologische Typen. Zürich: Rascher; 1921. English translation: Psychological Types. London: Kegan Paul; 1923.
Myers IB, McCaulley MH. Manual: A Guide to the Development and Use of the Myers-Briggs Type Indicator. Palo Alto, CA: Consulting Psychologists Press; 1985.
Pittenger DJ. Cautionary comments regarding the Myers-Briggs Type Indicator. Consulting Psychology Journal: Practice and Research 2005;57(3):210-221. doi:10.1037/1065-9293.57.3.210
The Myers & Briggs Foundation. Legal and trademark notices. https://www.myersbriggs.org/legal/

What can make this go out of date

None at runtime — static editorial pages.

16-Type Compatibility Matrix