Why You Keep Testing People You Love (And How It Pushes Them Away)

You pick a fight you didn't need to pick. You go quiet and wait to see if they'll notice. You drop a hint about something that's bothering you — then tell yourself it doesn't matter — then feel wounded when they don't pick up on it. You say you're fine when you're not. You create distance, then watch to see if they'll close it. You're not trying to be difficult. You're running a test. And the terrifying part is: you're not even fully aware you're doing it.

Relationship testing is one of the most common — and least talked about — patterns in adult intimacy. It's not manipulative in the way we usually mean that word. It's not calculated. It's desperate. It's a nervous system that learned long ago that love is conditional, inconsistent, or likely to disappear — and now it's constantly scanning for evidence. The test isn't 'do they love me.' The test is 'when will they stop.'

Psychologists define relationship testing as a set of behaviors designed to gather information about a partner's commitment, affection, or reliability — without directly asking for it. The key word there is 'without.' Because direct asking feels too vulnerable. If you ask and the answer disappoints you, you have to sit with that. If you test, you can tell yourself the results don't count.

Research by Leanne Knobloch and Denise Solomon on 'relational turbulence theory' shows that people engage most heavily in indirect testing when they're in a state of relational uncertainty — when they can't accurately predict a partner's behavior or feelings. The nervous system hates uncertainty. So it manufactures information. It creates scenarios to force a response. It interprets that response as data. The problem is the experiment is almost always rigged.

• ◆Indirect suggestion testingDropping hints about a need or concern to see if the partner notices without being told directly
• ◆Jealousy inductionMentioning attraction to others or flirting elsewhere to gauge the partner's reaction and confirm investment
• ◆Withdrawal testingGoing cold or distant to see if the partner will chase, reach out, or demonstrate care
• ◆Compliance testingMaking a request (or implying one) not because you need the thing, but to see if they'll prioritize you
• ◆Endurance testingCreating conflict, expressing difficult emotions, or behaving badly — then watching whether the partner stays

Testing doesn't emerge from nowhere. It's learned. John Bowlby's attachment theory — and decades of subsequent research by Mary Main, Mary Ainsworth, and others — established that children develop internal working models of relationships based on early caregiver experience. If your primary caregivers were inconsistent — warm one day, withdrawn the next, attentive when you performed and absent when you needed — your nervous system didn't learn 'love is safe.' It learned 'love is unpredictable, and I need to keep checking.'

This maps directly onto what we now call anxious attachment — the attachment style most closely associated with testing behavior. Anxiously attached adults didn't experience reliable emotional attunement in early life. They learned that expressing needs directly sometimes worked and sometimes didn't, so they developed indirect strategies: escalating emotion to get a response, withdrawing to see if the other person would notice, creating drama as a proximity-seeking behavior. These strategies were adaptive then. They kept connection alive in an unpredictable environment. In adult relationships, they corrode the very thing you're trying to protect.

What makes this more complex is that disorganized attachment — formed in environments where caregivers were also sources of fear or chaos — produces a particularly painful testing pattern. You desperately want closeness and are terrified of it at the same time. You test, the person passes, you feel momentarily safe, and then the fear returns and you test again. It's not a cycle you chose. It's a survival strategy that got locked in.

Here's what makes relationship testing so psychologically insidious: the tests are almost impossible to pass — not because your partner is failing, but because you've built in a structural flaw. When you test someone indirectly, you're not evaluating their actual feelings for you. You're evaluating their ability to read your mind, respond to signals they may not know exist, and demonstrate love in the specific way your nervous system requires at that exact moment.

Suppose you go quiet because you're hurt, and you're testing whether they'll notice. If they don't notice, you take it as confirmation: 'They don't care.' If they do notice, there's often a secondary test — is their response the right kind of noticing? Is the tone right? Did they ask the right question? Did they seem worried enough? Testing doesn't end when someone passes. It escalates. Each passed test raises the bar for the next one, because the underlying anxiety hasn't been addressed — only temporarily soothed.

Research by Jeffry Simpson and colleagues on stress and attachment found that anxiously attached individuals are primed to detect threats to the relationship even in ambiguous situations — and to interpret ambiguous partner behavior negatively. So your partner's distracted look during dinner isn't a neutral event. It's evidence. It confirms what part of you already believed. The test was never really a test. It was a verdict you were looking for confirmation to deliver.

If you've ever watched a relationship deteriorate without being able to stop it — even though you loved the person, even though they seemed to love you — relationship testing may be part of the mechanism. Securely attached and emotionally available partners can usually tolerate one or two tests. They sense something is off, they lean in, they try to connect. But sustained, repeated, escalating testing eventually produces something that looks a lot like the feared abandonment: exhaustion, withdrawal, resentment.

There's a clinical concept called 'coercive cycles' — patterns where one partner's anxiety-driven behavior triggers a response in the other partner that confirms the anxious partner's worst fears. You go cold to test them. They give you space because they've learned that's what you want when you go quiet. You interpret the space as rejection. You escalate. They pull back further. Now you have evidence. The abandonment you always knew was coming has finally arrived — and you created the conditions for it.

This isn't a character flaw. It's a trauma response meeting an environment that can't absorb it indefinitely. The tragedy is that the people most desperate for proof of love are using strategies that destroy love's ability to survive.

Testing feels like communication. It feels like you're expressing something — hurt, need, uncertainty. But it's the opposite of communication. Communication reduces relational uncertainty by providing actual information. Testing maintains relational uncertainty by keeping information hidden and watching what happens in the gap.

Real communication sounds like: 'I felt disconnected from you this week and I'm not sure why — can we talk about it?' Testing sounds like: saying nothing, being noticeably colder, waiting. The first is vulnerable. The second feels safer because if they don't respond, you can pretend you didn't really put anything on the line. But safety purchased through ambiguity is not safety. It's a holding pattern. You never learn whether this person can actually meet you — because you never let them see you clearly enough to try.

Developmental psychologist Diana Fosha's work on 'transformance' — the innate drive toward authentic emotional experience — suggests that what people actually want in relationships isn't the outcome of the test. It's the experience of being fully known and not abandoned. Testing is an attempt to get that experience without the risk. But the experience is only available through the risk.

The antidote to testing isn't willpower or a decision to 'stop being insecure.' That framing misses the point entirely. The behavior is downstream of a nervous system that was never given reliable evidence that direct vulnerability is safe. You can't logic your way out of that. What you can do is begin building new evidence — slowly, deliberately, and with increasing tolerance for the discomfort that comes before the safety.

The three practical moves that actually shift this pattern over time:

• ◆Name the urge before you act on itWhen you feel the pull to go cold, create drama, or drop a hint instead of speaking directly, pause and name it internally: 'I'm about to run a test.' That small moment of metacognition interrupts the automaticity of the pattern
• ◆Replace the test with the vulnerable statementInstead of manufacturing a situation to force a response, say the actual thing. 'I've been feeling insecure about us lately' is terrifying precisely because it's real. That terror is the signal you're doing it right
• ◆Audit what you're doing with the responseIf your partner responds well to direct vulnerability and you immediately find a reason why it doesn't count, or start preparing for the next test, the problem isn't your partner. The work is internal
• ◆Understand your baselineLearn your specific attachment wiring. Testing looks different in anxious, disorganized, and fearful-avoidant styles. What triggers it for you, what the test typically looks like, and what outcome would actually feel like enough are all questions worth sitting with seriously
• ◆Work the nervous system, not just the behaviorInsight alone rarely rewires deep relational patterns. The nervous system needs new emotional experiences — including experiences of being safe after being direct — to update its predictions about what intimacy means

Something genuinely different becomes available when you stop testing and start communicating. You get real information. You find out whether this person can actually meet your needs — not the performed, cryptically-signaled, indirect version of your needs, but the real ones. Some relationships don't survive that clarity. That's painful, but it's useful pain. The relationship was built on a dynamic, not a connection.

Most people who do this work are surprised to discover something else: they were far more lovable to the people around them than they knew. The testing behavior was filtering out the evidence. Partners who had been trying to reach them for months — years — suddenly became visible. Not because the partners changed. Because the person finally let them close enough to matter.

Intimacy isn't the absence of fear. It's the decision to be known anyway. That decision gets easier every time you make it and survive it. The nervous system updates slowly — but it does update. What was once proof that love was about to disappear can gradually become proof that it was there all along.