Friendship does shape personality, but the effect is smaller than most people assume, and it operates differently depending on your age. The strongest influence happens during adolescence, when friends can nudge your behaviors, emotional patterns, and social skills in measurable ways. By adulthood, the effect shrinks considerably. A large meta-analysis covering 60 studies across four continents found that the overall peer influence effect on behavior over time is statistically significant but small, with a standardized coefficient of .08.
What makes this topic tricky is that friends often seem to influence each other more than they actually do. Much of the similarity between friends comes not from changing each other, but from choosing each other in the first place.
Selection vs. Socialization: Why Friends Look Alike
Personality researchers distinguish between two forces that explain why friends tend to share traits and behaviors. Selection is the tendency to befriend people who are already similar to you. Socialization is the process of actually becoming more like your friends over time. Both are real, but research consistently shows that selection does more of the heavy lifting. A study from Washington University in St. Louis used advanced statistical methods to tease the two apart and found that when selection bias was properly controlled for, nearly all socialization effects disappeared.
This matters because it reframes the question. If you and your closest friend are both highly organized and dependable, it’s more likely you gravitated toward each other because of those shared traits than that one of you “taught” the other to be that way. That said, socialization effects aren’t zero, especially earlier in life.
How Friends Shape You in Adolescence
The teenage years are when friendship exerts its strongest pull on personality development. A review of longitudinal research found that socialization effects through friendships are moderately sized during adolescence, then drop to small or negligible levels from young adulthood onward. This lines up with what developmental psychologists have long suspected: adolescence is a sensitive period for identity formation, and friends play a central role in that process.
Friendships during this period work as a kind of training ground for social and emotional skills. When teenagers navigate the give-and-take of close relationships, they practice behaviors like offering support, asking for help, and managing conflict. Over time, these repeated experiences can shape lasting tendencies. Research published in Developmental Psychology found that high-quality friendships predicted decreases in avoidant behaviors during both help-giving and help-seeking tasks. In other words, teens with good friends got better at showing up for people and accepting support from them.
Friends also function as a feedback system. Social learning theory explains that friends reinforce certain behaviors, responding positively to kindness and cooperation while pushing back (sometimes subtly, sometimes not) against hostility or selfishness. Having multiple friends broadens this effect by exposing a young person to a wider range of social norms, helping them recognize which behaviors are typical in healthy relationships and which are not.
Adolescents with higher levels of psychological maturity (what researchers call ego development) tend to form deeper, more intimate friendships over time. One longitudinal study found that ego development predicted increases in intimate behavior with a best friend, stronger peer attachment, and growing popularity, even after accounting for earlier levels of those same variables. The relationship appears to run in both directions: maturity attracts better friendships, and better friendships support further maturity.
What Changes and What Stays Stable
The Big Five personality traits (openness, conscientiousness, extraversion, agreeableness, and neuroticism) are moderately to highly stable across adulthood. Test-retest correlations range from .54 to .70 over shorter intervals and .31 to .45 over spans of 20 to 50 years. If you’re more extraverted than most people in your twenties, you’ll likely still be more extraverted than most people in your fifties. Friendship doesn’t override this basic continuity.
Where friendship seems to matter most is at the edges. It can amplify or soften existing tendencies rather than create entirely new ones. A conscientious teenager who befriends other conscientious peers may become slightly more so over time. A naturally agreeable person in a supportive friend group may develop even stronger cooperative instincts. Research has found that extraversion and conscientiousness in early adolescence predict higher friendship quality at age 19, and conscientiousness specifically predicts improvements in friendship quality from 19 to 23. The relationship between personality and friendship quality is genuinely reciprocal, each feeding the other in small increments.
One theory, called the Social Investment Principle, proposes that investing in social roles (including the role of “friend”) should drive personality maturation. The idea is that the expectations friends place on you push you to become more responsible, agreeable, and emotionally stable. However, when researchers tested this directly by measuring the personality levels expected by friends, partners, and supervisors, the results were only partially supportive. Traits under stronger social expectations did show higher average levels, but changes in those traits over time weren’t clearly linked to how strongly they were expected.
The Cost of Harmful Friendships
If supportive friendships gently encourage positive development, conflictual or toxic friendships can push in the opposite direction. Research on friendship victimization, which includes relational aggression, verbal abuse, and controlling behavior, shows clear links to increased depressive symptoms. In one study, greater friendship victimization predicted more depressive symptoms even after controlling for other forms of peer victimization like bullying, explaining an additional 5% of the variance in depression scores. All three subtypes of friendship victimization (relational, physical/verbal, and controlling) independently contributed to worse outcomes.
The personality implications are indirect but real. Greater internalizing symptoms like depression and anxiety have been linked to increased controlling behaviors in friendships, creating a feedback loop. A teen who becomes more anxious due to a toxic friendship may then engage in more controlling behavior themselves, which strains other relationships and reinforces the anxiety. Over time, these patterns can settle into more trait-like emotional tendencies, particularly increased neuroticism. Having even one supportive best friendship, however, can buffer against the damaging effects of broader peer rejection.
Why the Effect Fades in Adulthood
By young adulthood, the window for friendship-driven personality change narrows significantly. Longitudinal research shows that selection effects on friendships intensify during adolescence, peak in young adulthood, and then diminish through middle and late adulthood. Socialization effects follow a different curve: moderate in adolescence, then small to negligible from young adulthood on.
Several factors explain this shift. Adults have more crystallized identities and are less susceptible to social pressure. They also tend to have fewer close friendships and less daily contact with friends compared to teenagers, who spend hours together in school and social settings. The meta-analysis on peer influence found that shorter time lags between assessments and peer contexts outside the classroom were associated with stronger influence effects, suggesting that sheer exposure matters. Adults simply have less of it with friends.
This doesn’t mean adult friendships are personality-neutral. They still contribute to well-being, emotional regulation, and life satisfaction. But the expectation that a new friend group in your thirties or forties will fundamentally reshape who you are isn’t well supported by the data. By that point, you’re mostly choosing friends who fit the person you’ve already become.