(FORTUNE Magazine) - Two heads are better than one. But what about four heads? Or ten? Why not 25? Picking the right size for a team depends on what the team needs to do. But researchers have also found that it is possible for a team to be too small. "We suggest you can't have a team of less than three because then it becomes a dyad," says Margaret Neale of Stanford Business School. A dyad is a partnership, not a team; adding a third person often raises productivity.

In 1972 psychologist Ivan Steiner examined what happens each time you increase the size of a team. Starting with one member and working his way up, he found that each additional member increased the group's potential productivity but also introduced inefficiencies: People became less motivated, and group coordination became more difficult. Steiner showed that team productivity peaked at about five. After that point, adding more members actually began to hurt the team's performance.

J. Richard Hackman, a professor of social and organizational psychology at Harvard, bans his students from forming project groups larger than six. Add just one more person, he writes in Leading Teams, "and the difference in how well groups of the two sizes operate is noticeable."

In 1970, Hackman and Neil Vidmar set out to find the perfect size. They asked teams large and small to do several tasks, then asked the participants how strongly they agreed with two statements: One, their group was too small for the task, and two, their group was too large. The optimum number: 4.6.

That's a theoretical number if there ever was one - though the Fab Four and their manager are an interesting real-life approximation. But the rule of 4.6 is still a good guideline, because research shows that teams tend to be overstaffed, often because managers don't want to leave anyone out.

It's a kind impulse, but too-large groups usually backfire. "If you don't break it down, people will break it down for you. They'll form cliques," says Jennifer Mueller, a professor at Wharton. Jeff Bezos has instituted a "two-pizza rule" at Amazon.com. If a team can't be fed by two pizzas, it's too large. The rule is hardly scientific, but it gets the point across: Keep it small enough for everyone to have a piece of the action.