Zero looks like the most obvious number there is — the natural starting point, the thing you get when nothing is left. But zero as a genuine, computable number, distinct from a mere empty placeholder, is one of the latest major mathematical inventions in the history of counting, not one of the earliest.
Two separate jobs had to be solved before zero became what it is today. The first is zero as a placeholder — a symbol marking an empty position in a positional number system, distinguishing 105 from 15 or 1,050. The Babylonians solved this by roughly the 3rd century BCE with a double-wedge placeholder mark within their base-60 system, though they never used it at the end of a number and it functioned purely as a spacer. Independently, the Maya developed a genuine placeholder zero (often shell-shaped) within their base-20 calendar system, likely by the 4th century CE or earlier — a wholly separate invention, with no contact between the two civilizations.
The second, much harder job is zero as an actual number you can compute with — add it, subtract it, and reason about it the way you would 5 or 12. This step is credited primarily to Indian mathematics. Brahmagupta, writing in 628 CE, laid out formal rules for zero's arithmetic: that a number minus itself gives zero, that a number plus zero is unchanged, and he even attempted (not entirely successfully) to define division involving zero — a problem still formally unresolved today, since dividing by zero has no meaningful answer within ordinary arithmetic. This built on a long Indian philosophical tradition already comfortable with concepts of emptiness and void, giving zero a conceptual foundation as much as a computational one.
From India, zero and the full positional number system spread through the Islamic world by the 8th and 9th centuries — mathematicians like Al-Khwarizmi (whose name gave us the word "algorithm") carried the system forward — and reached Europe gradually afterward, popularized in large part by Leonardo of Pisa's 1202 book Liber Abaci. The transition wasn't smooth: Roman numerals, with no zero and no positional structure, had been Europe's establishment system for over a thousand years, and some cities went so far as to restrict the new "Arabic" numerals in official records, partly from a fear that zero made numerical fraud easier to conceal.
Ancient Greek mathematics, remarkably advanced in geometry, largely avoided treating zero as a proper number — a reflection of a philosophical tradition that tied "number" closely to counting things that actually exist, and it's hard to count a collection of nothing. That's part of why zero's full arrival took so long: it wasn't simply a missing symbol waiting to be drawn, but a genuinely difficult conceptual leap that took multiple civilizations working somewhat independently to complete.
Modern zero does considerably more mathematical work than either of its ancient predecessors ever did. It's the additive identity (any number plus zero returns that number unchanged, the direct counterpart to how 1 functions as the multiplicative identity), the boundary separating positive numbers from negative ones on the number line, and the origin point every coordinate system in geometry and physics is built around. Division by zero remains formally undefined — not because mathematicians haven't gotten around to defining it, but because no consistent value could be assigned to it without breaking other, more fundamental arithmetic rules, a genuine mathematical dead end rather than an oversight.
It's also worth noting what zero is not, since the surrounding confusion is common: zero is an even number (it satisfies the standard definition of evenness, being evenly divisible by 2, with a quotient of exactly 0), it is neither positive nor negative, and — unlike every positive integer covered elsewhere on this site — it has no meaningful "factors" in the conventional trial-division sense, since every nonzero number technically divides evenly into it (0 divided by any nonzero number is 0), which would make a conventional factor list infinite and therefore not useful the way it is for any other integer.
On this site, zero sits in a category of its own among the computed number facts: it has no factors in the conventional sense, it's neither prime nor composite, and it isn't classified as a perfect, triangular, or Fibonacci number the way most other small integers are — not because it's excluded arbitrarily, but because those classifications are built around properties (like proper divisors summing to the number) that simply don't apply the same way at zero. It's a fitting reminder that even the "start" of the number line required real, hard-won mathematical work to define properly.