How this calculation works
Each distinct member receives a three-bit membership label. The seven nonzero labels partition the union into only A, only B, only C, three exclusive pair intersections and the triple intersection.
|A∪B∪C| = |A|+|B|+|C|−|A∩B|−|A∩C|−|B∩C|+|A∩B∩C|A worked example
A={apple,banana,cherry}, B={banana,cherry,date}, C={cherry,date,elderberry} have five union members. Only cherry belongs to all three.
Good to know
At most 200 entries per input set, each up to 80 characters. Text identity is case-sensitive after trimming and Unicode normalization; numeric spellings remain distinct labels. No outside count without a universal set. Diagram circles are schematic rather than proportional.
Assign each item to exactly one region
A three-set diagram has seven internal membership regions. An item can belong to just one set, exactly two sets, or all three. The phrase A intersection B includes the triple intersection, while the exclusive A-and-B region excludes C. Keeping those two counts distinct prevents double counting.
With the fruit example, banana belongs to A and B but not C. Cherry belongs to all three. Date belongs to B and C but not A. The member table exposes these assignments directly rather than requiring you to infer them from circle areas.
Use inclusion–exclusion as a check
Adding the sizes of A, B and C counts a member twice when it belongs to two inputs and three times when it belongs to all three. Subtract the three pair-intersection sizes, then add the triple intersection once. The result is the size of the union.
The seven disjoint region counts provide another route to the same total: simply add them. Their sum should equal the number of distinct labels across your lists. Duplicate labels in one input do not increase membership counts, because a set records presence rather than frequency.
Treat words as labels rather than quantities
Use commas or newlines to separate entries. Leading and trailing spaces are removed, and Unicode-normalized equivalents are merged. Other text is case-sensitive: Apple and apple remain different members. The labels 1 and 1.0 also remain distinct in this tool.
These conventions make the treatment of names and identifiers predictable. Do not paste a quoted CSV field that itself contains a comma, because the comma is still a separator. A multiset or frequency analysis requires a different data model. Use the set operations calculator when comparing two sets algebraically.
Read schematic circles honestly
The circles are arranged to show every possible overlap clearly, not to make their areas proportional to the counts. A large count in a small visual region does not indicate a mistake. The printed counts and member table are the quantitative outputs.
The outside region cannot be calculated from A, B and C alone. It needs an explicitly defined universe, such as all students in a class. If a survey has 30 participants and 24 belong to the union, then six are outside, but those extra six cannot be inferred from three selected lists alone.
A few questions, answered
Can I enter an empty set?
Yes. Enter {} or ∅.
Does repeating a label increase its count?
No. Each set records membership once.
Are the circles proportional?
No. They are schematic. Use the counts and member table.