The Thomas-Kilmann Conflict Mode Instrument, usually shortened to TKI, is a psychometric tool developed in the early 1970s by Kenneth Thomas and Ralph Kilmann to map how people handle disagreement. It doesn't measure aggression or passivity as traits — it maps behaviour on two dimensions: how assertive you are (pushing for your own interests) and how cooperative you are (attending to others' interests). From those two axes, five distinct conflict modes emerge, each with genuine strengths and real limitations depending on context.
The Two-Dimensional Framework
Thomas and Kilmann drew on the managerial grid work of Robert Blake and Jane Mouton, adapting it specifically for conflict situations. Their insight was that how you handle conflict can't be reduced to a single spectrum from "aggressive" to "passive." Two independent dimensions are at work simultaneously:
Assertiveness — the degree to which you try to satisfy your own concerns in a conflict. High assertiveness means pushing your position; low means letting it go.
Cooperativeness — the degree to which you try to satisfy the other person's concerns. High cooperativeness means actively working toward the other side's interests; low means focusing on your own.
These dimensions are independent, which is the key point. You can be highly assertive and highly cooperative simultaneously — that's the collaborating mode. Or highly assertive and uncooperative — competing. The combination creates the five modes.
The Five Conflict Modes
Each mode sits at a different position in the assertiveness-cooperativeness space:
- Competing (high assertiveness, low cooperativeness) — pursuing your position firmly, without significant attention to the other side. Appropriate when a decision needs to be made quickly, when you have higher authority or expertise, or when an issue is genuinely non-negotiable. Overused, it damages trust and shuts down information flow.
- Accommodating (low assertiveness, high cooperativeness) — conceding to the other side's position. Appropriate when the issue matters more to them than to you, or when preserving the relationship is the priority. Overused, it signals that your concerns don't matter and can build hidden resentment.
- Avoiding (low assertiveness, low cooperativeness) — sidestepping or postponing the conflict. Appropriate when the issue is trivial, when you need more time to gather information, or when the moment is genuinely wrong. Overused, it leaves problems to fester.
- Collaborating (high assertiveness, high cooperativeness) — working together to find a solution that fully addresses both parties' concerns. Appropriate for complex issues where both sets of interests matter and a creative solution is achievable. Requires significant time and goodwill; not always practical.
- Compromising (moderate assertiveness, moderate cooperativeness) — finding middle ground where both sides give something up. Faster than collaborating, and sometimes the right call when a good-enough solution is needed quickly. The risk is producing solutions that satisfy no one fully.
Why Situational Fit Matters
Thomas and Kilmann were emphatic that no single mode is universally superior. The TKI explicitly asks respondents to think about their typical behaviour, then provides feedback on whether their modal tendencies fit the situations they actually face. Someone who over-relies on avoiding in a technical role where disagreements need to be resolved will produce poor outcomes regardless of how "nice" that approach feels.
The most people have one or two preferred modes and one or two modes they almost never use. The unused modes are often the most interesting — they represent repertoire gaps that become costly in specific contexts. A manager who never competes will struggle in crisis decisions; one who never accommodates will gradually exhaust their team.
Using the Model Without the Instrument
The TKI is a scored instrument, but the conceptual framework is useful even without formal assessment. A few practical applications:
In team settings, naming the modes gives teams a shared vocabulary. "We've been avoiding this for three weeks" is easier to act on than a vague sense that something isn't being discussed.
In one-on-one conflict, identifying which mode each person is currently in can clarify why the conversation is stuck. Two people competing produce a fight. One competing and one accommodating produces a lopsided outcome. Two people avoiding produces nothing — the issue stays alive underground.
In performance conversations, managers who understand the model can calibrate: not "be nicer" or "be tougher," but "is the competing mode appropriate here, or is this a collaborative conversation?"
Limitations of the Framework
A few things the TKI doesn't capture well:
It describes behavioural tendencies, not fixed personality traits. The same person can use different modes in different relationships or contexts — the instrument captures typical behaviour, not essential character.
It doesn't account for power dynamics. The "appropriate" mode depends partly on relative power, authority, and stakes — none of which the basic model addresses.
The five modes sit neatly in a grid, but real conflicts are messier. People often shift modes mid-conversation. The framework is a diagnostic, not a script.
It also doesn't tell you how to execute each mode well. Competing can look like clear boundary-setting or like aggression, depending on how it's done. The model maps what mode you're in; skill determines how well you execute it.
The TKI in Organisational Research
Thomas-Kilmann data from thousands of organisations suggests that most managers over-use competing and avoiding relative to their ideal profile. Collaborating is under-used, particularly in situations where it would produce significantly better outcomes — though the model acknowledges this may be partly realistic: genuine collaboration requires time and conditions that aren't always available.
The pattern shifts with seniority and organisational culture. Technical and operational roles tend to show higher competing scores; roles requiring stakeholder management tend to show higher accommodating. Neither profile is wrong; the question is whether it fits the actual demands of the role.
To see where your own tendencies sit across conflict and related interpersonal styles, our free conflict styles assessment maps your profile across the five Thomas-Kilmann modes and provides detailed feedback on which contexts suit each.
Frequently Asked Questions
What are the five Thomas-Kilmann conflict modes?
Competing (high assertiveness, low cooperativeness), Accommodating (low assertiveness, high cooperativeness), Avoiding (low on both), Collaborating (high on both), and Compromising (moderate on both). Each has appropriate use cases and costs when over-used.
Is one conflict mode better than the others?
No. Thomas and Kilmann explicitly argued that all five modes are useful in different situations. The question is whether your habitual modes match the contexts you face. Most people over-rely on two or three modes and under-use others — the gaps become costly in specific situations.
How is the TKI different from other conflict models?
The TKI's distinctive contribution is the two-dimensional framework: assertiveness and cooperativeness as independent variables. This produces five distinct modes rather than a single spectrum, and it separates the question of "how hard am I pushing" from "how much am I attending to the other side."
Can you change your conflict mode?
Yes. The TKI measures behavioural tendencies, not fixed traits. With awareness and practice, people expand their repertoire — learning to collaborate when they default to avoiding, or to accommodate when they default to competing. The instrument is designed to support that kind of deliberate development.
Is compromising the same as collaborating?
No. Compromising is faster and involves both sides giving something up to reach a middle position. Collaborating takes longer but aims for a solution that fully satisfies both sides' core interests — no sacrifices needed, but more effort and goodwill required. Which is appropriate depends on time available, relationship stakes, and whether a genuinely creative solution exists.