In short
Division by zero is undefined — and not because mathematicians are being fussy. Algebraically, 5 / 0 would have to be a number x such that 0 \cdot x = 5, but 0 \cdot x is always 0, never 5. Geometrically, division is the slope of a line — and dividing by zero is the slope of a vertical line, which has no finite value. The practical danger: when you solve an equation by dividing both sides by an expression in x, that expression might be zero — and you will silently lose a solution without noticing.
You have done it a hundred times. You punch 5 \div 0 into your calculator just to see what happens, and the screen flashes Error or Undefined. Your teacher mutters "you can't divide by zero" and moves on. But why? Zero is just a number. You can multiply by it (you get zero), you can add it (you get the same thing back), you can subtract it. Why does division alone refuse to play?
The answer is not a rule. It is a consequence — of what division means, of what slopes look like, and of what happens to equations when you are careless. Once you see the picture, the rule stops being arbitrary.
The algebraic proof — division is the inverse of multiplication
Here is the secret nobody tells you in class 6: division is not really an operation. It is a question. When you write a / b, you are asking, "What number x multiplied by b gives a?" In symbols:
Why: this is just the definition. 10 / 2 = 5 because 2 \cdot 5 = 10. 21 / 3 = 7 because 3 \cdot 7 = 21. Division is multiplication run backwards.
Now plug in b = 0. The question 5 / 0 = x becomes:
What value of x makes this true? None. Whatever real number you pick — 1, -7, \pi, a billion — multiplying it by 0 gives 0, never 5. The question has no answer, so 5/0 has no value. We say it is undefined.
Now try the trickier case: 0 / 0. The question becomes:
This time every real number x works. x = 1? Yes, 0 \cdot 1 = 0. x = -42? Yes. x = \sqrt{2}? Yes. The question has infinitely many answers, so it has no unique answer. Mathematicians refuse to assign one value when there is no way to pick fairly. So 0/0 is also undefined — for the opposite reason. One has no answers, the other has too many.
The calculator test
Open the calculator app on your phone. Type 5 / 0 and press equals. You will see "Error", "Cannot divide by zero", or "Infinity" depending on the app. None of those is a number you can do further arithmetic with. Now type 0 / 0. Same story. The hardware itself refuses, because the chip designers built in the same algebraic check you just did — there is no x to return.
The geometric picture — slope of a vertical line
Now switch from arithmetic to geometry. Plot two points on a graph and draw the line joining them. The slope is "rise over run":
Pick two points on a non-vertical line, say (1, 2) and (4, 8). Slope is (8-2)/(4-1) = 6/3 = 2. Fine. Now pick two points on a vertical line, say (3, 1) and (3, 7). Slope is (7-1)/(3-3) = 6/0. Undefined.
The blue line goes up and across — both rise and run are positive, slope is a finite number. The red line goes straight up — rise is real, but run is zero. There is no number you can multiply 0 by to get the rise, so the slope simply does not exist as a real number.
This is the same thing as the algebraic argument, dressed up in geometry. "Slope = 6/0" is asking, "How much does y change for each unit x moves?" — but x never moves on a vertical line. The question is incoherent.
How $1/h$ explodes as $h$ shrinks
Compute 1/h for a sequence of h values getting closer to zero:
| h | 1/h |
|---|---|
| 1 | 1 |
| 0.1 | 10 |
| 0.01 | 100 |
| 0.001 | 1{,}000 |
| 0.0001 | 10{,}000 |
The numbers blow up without bound. There is no real number that 1/h "approaches". This is why calculus says \lim_{h \to 0} 1/h does not exist — the spike has no top.
This picture is the bridge to calculus. When functions explode like this near a point, that point is called an asymptote. The graph of y = 1/x has a vertical asymptote at x = 0 for exactly the same reason — 1/0 has no value, but the curve gets infinitely tall as x approaches 0. Every later topic about limits, continuity, and asymptotes traces back to this one fact.
Why this matters when solving equations — the lost-solution trap
Now to the dangerous part. When you solve linear equations, you constantly divide both sides by something. Usually that something is a clean nonzero constant — divide by 7, divide by -3. Safe. But what if the thing you divide by is an expression in x? Then it might be zero, and you will not always notice.
Consider this innocent-looking equation:
You might think, "Easy, divide both sides by x":
Done? Not quite. Plug x = 0 back into the original: 0^2 = 3 \cdot 0, that is, 0 = 0. True. So x = 0 is also a solution — but your method silently discarded it, because dividing by x is dividing by zero when x = 0. The original equation had two solutions, x = 0 and x = 3, and you only kept one.
Why: dividing by x assumes x \neq 0. That assumption throws away the case x = 0 before you ever get to check it. The safe move is to bring everything to one side and factor: x^2 - 3x = 0 \Rightarrow x(x-3) = 0 \Rightarrow x = 0 or x = 3. Both solutions survive.
The $x^2 = 3x$ trap, fully unpacked
Wrong route — divide both sides by x:
Right route — move everything to one side, factor:
A product is zero exactly when one of the factors is zero, so x = 0 or x - 3 = 0, giving x = 0 or x = 3. Both values check in the original equation. Two solutions, not one.
The same trap appears in equations with fractions. If you cancel a factor like (x - 2) from numerator and denominator, you are dividing by x - 2, which equals zero at x = 2. That value might have been a real solution. Always note any value where a denominator vanishes before you cancel.
This is one face of the broader topic of extraneous solutions — except in reverse. There you gain false solutions; here you lose true ones. Both come from the same root cause: doing arithmetic that quietly assumes something is not zero.
The takeaway
Three pictures of the same fact. Algebraically, a/0 is the question "what times zero gives a?" — and that question has no honest answer. Geometrically, division is slope, and the slope of a vertical line points nowhere. Practically, in equation-solving, dividing by an unknown that could be zero erases solutions without warning.
Treat any division by an expression in x with suspicion. Ask: "Could this be zero?" If yes, factor instead of cancelling, and check the suspicious value separately. Your equations will keep all the solutions they deserve.