Quick answer
Jeffrey Hall's studies put casual friendship at roughly 40 to 60 hours together, friendship at about 80 to 100, and close friendship beyond 200. Hours spent only working side by side count for less, and both people have to want the time.
Platonic closeness is largely a time-and-activity problem. Jeffrey Hall's studies at the University of Kansas estimated how many hours of time together separate an acquaintance from a casual friend, a casual friend from a friend, and a friend from a close friend. The estimates are ranges, they differ for new students and for adults, and hours spent merely working beside someone count for less than hours you chose. A separate line of public-health research finds that people with stronger social relationships are more likely to survive over time. That finding is about relationships in general, not a prize awarded to friendship as a concept, and it does not make a close friend appear because you wish for one.
The practices below come from those findings: invite someone, change the setting, and spend the kind of hours that actually move a bond.
The hour thresholds, with the ranges kept
Hall reported the work in the Journal of Social and Personal Relationships (the paper is dated 2019 in the journal's listing; the University of Kansas announced it in 2018). The KU summary, which Hall stood behind in interviews, is the figure most people can check without a paywall: roughly 50 hours together to go from acquaintance to casual friend, about 90 hours to reach "friend," and more than 200 hours before "close friend." Combining his two studies, he put the transitions at about 40 to 60 hours for a casual friendship, 80 to 100 hours to become a friend, and more than 200 hours together to become good friends.
The journal article is more precise, and the precision matters so you do not treat 50/90/200 as a law. In the student study, the chance of calling someone a casual friend rather than an acquaintance was greater than half after about 43 hours together in the first three weeks. Adults who had relocated, reporting on relationships that had already lasted months, needed more like 94 hours for that same casual step, a number Hall treats as inflated by the longer window. The paper's own combined reading is the one to remember: somewhere between 40 and 60 hours for a casual friendship in the early weeks, and a better-than-even chance of moving from casual friend to friend after about 80 to 100 hours. Students reached good or best friendship sooner than adults. Greater Good's write-up of the same studies puts the student good-friend threshold near 119 hours and notes that adults needed on the order of another hundred hours beyond that.
Who was studied is part of the result. One sample was 355 adults who had moved in the previous six months and were looking for friends in a new place. The other was 112 first-year students in Lawrence, Kansas, asked about two people they had met since arriving, then followed up. These are newcomers building new ties, not a census of friendships that have lasted twenty years. If your oldest friend would not survive a stopwatch audit of the last month, that does not refute Hall. His question was how a new bond crosses a threshold, not how a long bond is maintained once it exists.
Which hours count
Hall's blunt finding is the one that changes a calendar. Time together included hanging out, joking, and playing. "Hours spent working together just don't count as much." Sitting in the same office, the same class, or the same group chat project can produce familiarity without producing a friend. The investment that moved people up the scale was time in which they were actually with each other, not time in which a task kept them adjacent.
He also found that transitions are lumpy. "When people transition between stages, they'll double or triple the amount of time they spend with that other person in three weeks' time." Some first-year students spent about a third of their waking hours in a month with one good friend. That is not a recommendation. It is a description of how fast the hours accumulate when a bond is actually moving. A friendship that stays at "we should get coffee" for a year is not slowly ripening. In these data, it is failing to change the dose.
The other constraint is consent. Hall told KU, "You can't make people spend time with you, but you can invite them." Both people have to want the tie. An invitation that gets a real yes, followed by a second setting that is not the original task, is the mechanism he names: "If you work together, go to lunch or out for a drink. These things signal to people that you are interested in being friends with them." The signal is the point. Staying after a meeting to talk about the meeting is still work. Leaving the building for a meal is a bid of a different type. It can be declined, and the decline is information, not a reason to become more intense.
Hall's article also nods to earlier estimates that people maintain on the order of 150 friendships in total, a limit associated with Robin Dunbar's social-brain research. He invokes it as a constraint on attention, not as a target. Two hundred hours with one new close friend is a large share of a finite budget. Choosing that person means not choosing others. Platonic intimacy, in this research, is expensive on purpose.
What the health research does and does not say
Julianne Holt-Lunstad, Timothy Smith, and J. Bradley Layton's 2010 meta-analysis in PLOS Medicine pooled 148 prospective studies and 308,849 people, followed for an average of 7.5 years. The average effect was an odds ratio of 1.50 (95 percent confidence interval 1.42 to 1.59): a 50 percent increased likelihood of survival for people with stronger social relationships. The association was larger for complex measures of social integration (odds ratio 1.91) than for crude markers such as living alone versus with others. It held across age, sex, initial health, follow-up length, and cause of death in their checks. The authors compare the magnitude with quitting smoking and say it exceeds well-known risks such as obesity and physical inactivity.
Read the design before you turn it into a slogan about best friends. These were observational prospective studies. They support a predictive association between social relationships and mortality. They do not prove that adding one friend causes a longer life, and the authors say plainly that interventions need more research. They also note that most of the evidence concerns naturally occurring relationships, and that support provided by strangers in formal programs has a weaker record. Received support was less predictive than social integration. Being woven into a set of relationships predicted survival better than having someone assigned to help you.
The U.S. Department of Health and Human Services, summarizing Holt-Lunstad's 2015 follow-up on its social-connection fact cards, states a related and harsher cut: social isolation is associated with a 29 percent increase in the risk of premature mortality. A separate card, citing Valtorta and colleagues (2016), links poor relationships, isolation, and loneliness with a 29 percent higher risk of heart disease and a 32 percent higher risk of stroke. Those cards are about isolation and weak ties as health risks. They are not a claim that every deep friendship is medically required, and they are not a reason to cling to a harmful one. Holt-Lunstad's 2010 discussion even flags that some relationships are detrimental. The health case for platonic bonds is that integration matters on the order of risks medicine already respects. It is not that any particular friend is obligatory.
Practices that follow the evidence
Invite in a form that can be accepted or refused in one move. Hall's line is the whole tactic: you cannot force the hours, and you can offer them. A specific proposal ("Thursday, the noodle place, an hour, not about the project") is an offer of the kind of time his study counted. "We should hang out sometime" does not put an hour on the ledger.
Change the context on purpose when the only hours you have are task hours. Coworking, classmates, and committee seats create proximity and, in his results, underperform as friendship fuel. One meal, one walk, one game, repeated, is a different investment. The signal is that you are choosing the person rather than the assignment.
Budget the threshold you actually want. Forty to sixty hours is a casual friend, not a confidant. Eighty to one hundred is the paper's zone for "friend." Close friendship sat beyond two hundred in the combined estimate, and adults were slower than students. If you want a close tie with someone you met at work in March, the math is a standing hour most weeks for a long time, plus a jump in time when the relationship is ready to change stage. Waiting to "feel close" before spending the time reverses the finding. In these studies the hours came first.
Protect the budget. A 150-person outer circle and a 200-hour close friendship cannot both be pursued at full strength. Intimacy here means fewer people getting the non-work hours, not a warmer attitude toward everyone you already see in meetings. Say no to a third weekly obligation so the one friendship can double its time for a few weeks. Hall observed that doubling or tripling during a transition. You can plan a smaller version without pretending you are a first-year student with an empty calendar.
Use the health evidence as a reason to take the hours seriously, not as a reason to medicalize your social life. Integration predicted survival in Holt-Lunstad's pool. Isolation is associated with earlier death on the HHS cards. Neither result is satisfied by collecting contacts. Complex social integration did more in the 2010 analysis than a binary "lives with someone." A life with several people who actually know you is closer to what was measured than a life with one roommate and a hundred muted group chats.
Revisit new ties on a calendar, not on a vibe. The student study checked back after a few weeks. You can do the same: after a month, have you accumulated anything like the casual-friend range, or are you still trading work talk? If the other person declines the non-work invitations, stop adding intensity. The studies required two people. One-sided pursuit does not become platonic intimacy by lasting longer.
Sources
- Hall, J. A. (2019). How many hours does it take to make a friend? Journal of Social and Personal Relationships. https://doi.org/10.1177/0265407518761225 and https://journals.sagepub.com/doi/full/10.1177/0265407518761225
- University of Kansas News, 28 March 2018, "How to make friends? Study reveals how many hours it takes": https://news.ku.edu/news/article/2018/03/06/study-reveals-number-hours-it-takes-make-friend
- Greater Good Science Center summary of the same studies: https://greatergood.berkeley.edu/article/item/how_long_does_it_take_to_make_a_friend
- Holt-Lunstad, J., Smith, T. B., & Layton, J. B. (2010). Social relationships and mortality risk. PLOS Medicine, 7(7), e1000316. https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1000316
- U.S. Department of Health and Human Services, social-connection fact cards (isolation and the 2015 Holt-Lunstad estimate; heart disease and stroke via Valtorta et al., 2016): https://www.hhs.gov/surgeongeneral/reports-and-publications/connection/index.html
Frequently Asked Questions
- How many hours does it take to make a friend?
- Hall's combined estimate, reported by the University of Kansas and in the Journal of Social and Personal Relationships, is about 40 to 60 hours together for a casual friend, 80 to 100 hours for a friend, and more than 200 hours for a close friend. Students moved faster than recently relocated adults.
- Do hours at work count?
- Hall found that hours spent working together do not count as much as hanging out, joking, and playing. He suggests leaving the work setting, for lunch or a drink, because the change of context signals that you want a friendship.
- Did the mortality research prove that one new friend lengthens your life?
- Holt-Lunstad, Smith, and Layton's 2010 meta-analysis of 148 studies found a 50 percent higher likelihood of survival among people with stronger social relationships. That is a predictive association in observational studies, not proof that adding a single friend causes a longer life.
- What does isolation do to mortality risk?
- HHS social-connection fact cards, citing Holt-Lunstad and colleagues 2015, say social isolation increases the risk of premature mortality by 29 percent. A separate card cites higher heart-disease and stroke risk for poor relationships, isolation, and loneliness.
- What is a concrete way to start?
- Make one invitation the other person can accept or decline in a single reply, in a setting that is not the task where you met. Then look after a month at whether any real hours accumulated. If they keep declining, stop increasing the intensity.
- Who was in Hall's studies?
- Three hundred fifty-five adults who had moved within six months, and 112 University of Kansas first-year students followed over their first weeks. The estimates describe new ties, not the maintenance of a friendship that has already lasted decades.