We tend to describe friendship as chemistry. You meet somebody, the conversation moves easily, and something clicks.

That description is not wrong. It is simply incomplete. A person can seem promising in the first hour and still remain an acquaintance because neither of you builds a second, tenth or fiftieth hour around the first.

A study by University of Kansas communication researcher Jeffrey Hall tried to put numbers around that accumulation. Its best-known conclusion is that it takes roughly 40 to 60 hours together to move from acquaintance to casual friend, around 80 to 100 hours to become a friend, and more than 200 hours to become close.

I like the finding because it makes friendship look less like fate and more like something built. But the famous numbers need some unpacking. They are estimates drawn from two US samples, not universal deadlines.

The paper combined recent movers with first-year students

The 2019 paper, “How many hours does it take to make a friend?”, reported two studies.

In the first, 355 US adults who had moved at least 50 miles within the previous six months chose one person they had met since relocating. They estimated how long they had known that person, how many hours they typically spent together and what they did in that time.

This part was retrospective. The researchers did not watch those adult friendships develop. They asked people to reconstruct the hours after the fact.

The second study followed 112 first-year students who had recently moved to a Midwestern university. Each named two new acquaintances, excluding roommates, old friends, relatives and romantic partners. The students reported on those relationships three, six and nine weeks into the semester.

That gave the paper a longitudinal component. At the same time, only 62% of the original students completed all three waves, and the group was young and mostly white. That matters when people turn its estimates into a rule for friendship at every age.

The famous figures are rounded summaries

The clean sequence of 50, 90 and 200 hours did not come from three exact cut-offs in one analysis.

Among the students, the estimated probability of being classed as a casual friend rather than an acquaintance passed 50% at 43 hours. The casual-friend-to-friend threshold appeared at 57 hours, while friend to good or best friend passed 50% at 119 hours in that early wave.

The recent movers produced higher estimates: 94 hours, 164 hours and 219 hours for the same transitions. Hall argued those adult figures were probably inflated because hours kept accumulating after a relationship had already changed category.

Looking across both samples and earlier work, he settled on broader ranges: 40 to 60 hours for a casual friendship, 80 to 100 for friendship, and more than 200 for good or best friendship.

So “50, 90 and 200” is useful shorthand. It is not a stopwatch built into human relationships.

Time predicted change, but it did not prove cause

In the student sample, more time together predicted greater closeness even after the previous wave’s closeness was included in the analysis. Students whose relationships moved into a closer category had also tended to add more hours.

That is stronger evidence than a single snapshot, but it still does not prove that adding hours caused friendship. People usually choose to spend more time with those they already like. Compatibility can create time just as time can create opportunities for closeness.

An earlier longitudinal study of 84 first-year students reached a compatible but less numerical conclusion. The relationships that became close differed in both the quantity and intimacy of their interaction. As those friendships developed, intimacy explained more of their strength beyond sheer quantity.

That does not independently reproduce the 50, 90 or 200-hour thresholds. It supports the broader idea that repeated interaction matters while also warning us that hours alone are not the whole story.

Not every hour carried the same social meaning

In Hall’s adult sample, a greater share of time spent hanging out or watching films, television or gaming was associated with more closeness. Time spent together at work or in class was associated with less closeness once other factors were considered.

This is not evidence that work prevents friendship. It suggests that involuntary proximity and chosen leisure are not interchangeable. Seeing somebody across a desk for 200 hours does not necessarily communicate the same thing as choosing to spend a Saturday afternoon with them.

The student results also distinguished total talking from particular forms of talk. Catching up, playful talk, serious conversation and expressions of attention or affection predicted increases in closeness beyond hours together. Small talk later in the term was associated with decreases in closeness.

That last finding should be held lightly. The small-talk measure covered only four topics and did not show that ordinary conversation damages friendship. I have written before about small talk as a signal of goodwill. It may open a relationship without being enough to deepen it by itself.

The arithmetic explains why adult friendship can feel slow

Four hours together each week would take about three months to reach 50 hours, a little over five months to reach 90, and almost a year to reach 200. At two hours a week, those spans roughly double.

That arithmetic is not a prescription. It simply shows how a perfectly good potential friendship can remain thin when meetings are pleasant but rare.

Since becoming a father, I have noticed how easily social time gets treated as whatever survives after work and family logistics. The finding made me see that arrangement differently. If friendship needs accumulated, freely chosen time, then protecting none of it is not a neutral decision.

A large analysis of nine billion mobile calls made a related point from another direction. The researchers found that people with larger communication networks did not devote proportionally more time to them and had weaker average ties. The Oxford research record describes time as a limited resource distributed unevenly across a social network.

Chemistry may begin a friendship, but repetition gives it somewhere to go

The Hall paper cannot tell us that any two people will become close after hour 201. Its samples were American, its categories were self-reported, the adult estimates depended on memory, and it did not randomly assign people to spend more or less time together. I could not find a direct independent replication of its exact thresholds.

What it offers is a useful change of scale. We often evaluate a new friendship after two coffees and decide the chemistry was not strong enough. Yet two coffees may represent four hours inside a process that commonly takes dozens.

The less romantic truth is also, in its way, encouraging. Closeness does not always arrive as instant recognition. Sometimes it is the result of returning, lingering and choosing the same person again until familiarity has had enough time to become a shared life.