What Is Conflict Resolution? The Thomas Kilmann Model Explained

August 18, 2026

What Is Conflict Resolution? The Thomas Kilmann Model Explained

Tuesday, 18 August 2026

Conflict resolution is the process of addressing a disagreement so that the underlying issue is settled rather than suppressed. The Thomas Kilmann model describes five ways people respond, plotted on how far they push for their own concerns and how far they accommodate the other person’s.

Key takeaways

The five modes are competing, collaborating, compromising, avoiding and accommodating, mapped on assertiveness and cooperativeness.
No mode is correct in general. Each is right in some situations, and the skill is matching the mode to the circumstance.
The model’s co-creator argues organisational systems drive most conflict behaviour, with individual preference explaining far less.
In Australian and New Zealand work health and safety terms, conflict is often a symptom of poor work design rather than a personality problem.

What is conflict resolution?

Start with an example. Two managers disagree about which team owns a new reporting process. It surfaces in a meeting, someone says they will take it offline, and nobody does. Six weeks later the report was late, each manager assumed the other was producing it, and the conversation that follows is no longer about process ownership. It is about trust. The disagreement did not go away when it stopped being discussed. It changed form.

Conflict resolution is the work of preventing that shift. It means addressing a disagreement so the substantive issue is actually settled, rather than left to convert into a relationship problem that is much harder to solve. Resolution does not require agreement or harmony. It requires a decision that both parties understand, and enough of a working relationship left over to implement it.

That framing matters because most workplace advice treats conflict as something to be minimised. Disagreement about how work should be done is normal and often useful. What causes damage is disagreement that is never resolved, or that is resolved in a way that leaves one party feeling the process was unfair.

What is the Thomas Kilmann model?

Kenneth Thomas and Ralph Kilmann developed their model in the early 1970s, building on Blake and Mouton’s earlier managerial grid. It describes conflict behaviour along two dimensions: assertiveness, meaning how far you attempt to satisfy your own concerns, and cooperativeness, meaning how far you attempt to satisfy the other person’s. Those two axes produce five modes. Competing is assertive and uncooperative. Accommodating is cooperative and unassertive. Avoiding is low on both. Collaborating is high on both. Compromising sits in the middle of each.

The framework has lasted because it removes the moral loading from behaviour that usually attracts it. Avoiding is not cowardice, it is a choice to leave something alone. Competing is not aggression, it is prioritising your own concern over the relationship for this decision. Naming the modes lets a team discuss what happened without anyone defending their character.

The instrument itself is a thirty item questionnaire producing a profile across the five modes, widely used in leadership development. Its value in practice is less about the score than the vocabulary. Once a team knows the five words, the conversation shifts from who is being difficult to which mode this situation actually calls for.

When should you use each conflict mode?

The point of the model is that no mode is correct in general. Competing fits when a decision is genuinely urgent, when safety or compliance is at stake, or when you hold accountability that cannot be delegated by consensus. Accommodating fits when the issue matters far more to the other person than to you, or when you were wrong and saying so quickly is worth more than defending the position.

Avoiding fits when the issue is trivial, when emotions need time to settle, or when you have no standing to affect the outcome. Compromising fits when the parties have comparable power, the goals are incompatible and a workable answer now beats a perfect answer later. Collaborating fits when the issue is important enough to justify the time, when both sets of concerns must genuinely be met, and when the relationship needs to survive intact.

Two errors follow from misreading this. The first is treating collaboration as the answer to everything, which burns hours on decisions that did not warrant them and quietly signals that nothing is straightforward enough to just decide. The second is having only one mode available. Most people have a default that served them well somewhere and now runs automatically, and the useful development goal is access to the other four rather than a different favourite.

Is conflict at work always bad?

The common advice is that task conflict is healthy while personal conflict is damaging. The evidence is less comfortable than that. Carsten De Dreu and Laurie Weingart’s meta-analysis of research on team conflict found strong negative correlations between relationship conflict and both team performance and member satisfaction, and also found negative correlations for task conflict, contrary to what both the academic literature and introductory textbooks had predicted. The negative association was stronger for complex work such as decision making and project work than for routine production tasks.

That result deserves care rather than overstatement. Later research has qualified it, and the two conflict types are correlated in practice, which makes them hard to separate cleanly: an argument about the work often becomes an argument about the person. The defensible reading is not that disagreement should be suppressed, but that the popular claim that task conflict improves team performance is not well supported. What matters more is whether disagreement gets resolved and whether the process feels fair.

Why does the same conflict keep happening?

When the same dispute recurs with different people, the cause is usually structural rather than personal. SafeWork New South Wales makes this point directly in its guidance, advising that businesses consider other psychosocial hazards that may contribute to conflict, since high job demands or a lack of role clarity can lead to incivility between team members. It also recommends that conflict investigations identify the source of the conflict rather than only addressing the behaviour, which may mean changing how the work is done.

This is not a fringe view about workplaces. Ralph Kilmann himself argues that an organisation’s systems and processes affect around 80 per cent of how conflict is addressed and resolved at work, while individual preferences for particular modes account for at most about 20 per cent of observable behaviour. That is the co-creator of the instrument saying his own model captures the smaller share. The implication for anyone running conflict training is uncomfortable but useful: teaching five modes to people working inside a structure that rewards avoidance will not change much.

The practical version is to ask what the recurring dispute is telling you about the design of the work. Two teams that argue every quarter about the same handover usually have an undefined boundary rather than a personality clash. Fixing the boundary resolves it permanently. Mediating the personalities resolves it until next quarter.

What happens when conflict is left unresolved?

Unresolved workplace disputes have a formal destination, and the volume is rising. In its 2024 to 2025 annual report the Fair Work Commission recorded 44,075 lodgments, an increase of 24 per cent above the five year average, driven mainly by unfair dismissal and general protections dismissal applications, with the Commission’s President describing the rate of growth as unsustainable within current structures. Not every one of those matters began as an unaddressed disagreement, and it would be wrong to read the figure that way, but it does show the scale of what arrives once internal resolution has failed.

The escalation pattern inside an organisation is more consistent. A disagreement that is never addressed becomes an assumption about intent, the assumption becomes an account shared with colleagues, and by the time it reaches a formal complaint the original issue is often no longer the point. SafeWork New South Wales notes that conflict which continues becomes harder to address, harder to repair, and may develop into workplace bullying, which is why responding quickly and fairly matters more than responding perfectly.

The harder skill is not knowing which mode to select. It is running the conversation once you are in it, when the other person is defensive, the history is longer than the issue and you have to separate what happened from what you concluded about it. That is what AcademyGlobal (AG) built its Effective Conflict Communication Skills training around, working through the conversations participants are actually avoiding rather than tidy case studies.

The question worth asking of any recurring conflict is whether you are dealing with two people who need a conversation or a system that keeps producing the same argument. The answer changes what you should do next.

Frequently asked questions

What are the five conflict management styles?

Competing, collaborating, compromising, avoiding and accommodating. They are defined by how assertive you are in pursuing your own concerns and how cooperative you are in addressing the other person’s. Competing is high assertiveness and low cooperativeness, collaborating is high on both, and avoiding is low on both.

Which conflict style is best?

None in general, which is the point of the model. Each mode suits particular circumstances, so competence means having access to all five and reading which one the situation calls for. Someone who collaborates on everything is as limited as someone who avoids everything.

Is avoiding conflict ever the right choice?

Yes. Avoiding is appropriate when the issue is genuinely trivial, when tempers need time to cool before a useful conversation is possible, or when you have no ability to influence the outcome. It becomes a problem when it is a reflex rather than a decision, since the issue usually returns in a less manageable form.

How accurate is the Thomas Kilmann instrument?

It is a self report questionnaire measuring preference rather than a diagnostic test, and published analyses have reported moderate reliability. Treat the profile as a starting point for discussion rather than a fixed description, particularly since people behave differently in different settings and with different people.

How do you resolve conflict between two team members?

Address it early, speak to each person separately first, and separate what was observed from what was concluded about intent. Look for the work related cause, since recurring conflict often reflects unclear roles or competing demands rather than incompatible personalities. Where behaviour may amount to bullying or harassment, follow your formal process rather than mediating.