In short
The discriminant of a quadratic equation ax^2 + bx + c = 0 is the number D = b^2 - 4ac. Its sign alone decides the nature of the roots: D > 0 means two distinct real roots, D = 0 means one repeated real root, and D < 0 means two complex conjugate roots with no real solutions. Geometrically, D tells you whether the parabola cuts, touches, or misses the horizontal axis.
Solve x^2 - 6x + 9 = 0 and you get one root, x = 3. Solve x^2 - 6x + 8 = 0 and you get two, x = 2 and x = 4. Solve x^2 - 6x + 10 = 0 and you get none (at least, none that are real). Three equations that look almost identical -- the only thing that changed is the constant term -- yet they behave in completely different ways.
There is a single number hidden inside every quadratic that predicts which of these three worlds you will land in, before you solve. It is called the discriminant, and it is the expression under the square root in the quadratic formula: D = b^2 - 4ac. Compute it first, and you already know the shape of the answer.
Why the discriminant controls everything
Look at the formula again: x = \dfrac{-b \pm \sqrt{D}}{2a}. The square root \sqrt{D} is the fork in the road.
-
If D > 0, then \sqrt{D} is a positive real number. The \pm produces two different values — one with +\sqrt{D}, one with -\sqrt{D}. You get two distinct real roots.
-
If D = 0, then \sqrt{D} = 0. The \pm contributes nothing — both branches give the same answer, x = -b/(2a). You get one repeated real root (sometimes called a "double root").
-
If D < 0, then \sqrt{D} is not a real number. You need the imaginary unit i = \sqrt{-1} to write it: \sqrt{D} = i\sqrt{|D|}. The two roots become a complex conjugate pair, and there are no real solutions.
Three cases, three different worlds — and a single number decides which one you are in. That is why computing D first, before doing anything else, is standard practice. It tells you the shape of the answer before you reach it.
The three cases, geometrically
The equation ax^2 + bx + c = 0 asks: where does the parabola y = ax^2 + bx + c cross the x-axis? The discriminant answers this geometrically.
The vertex of the parabola y = ax^2 + bx + c sits at x = -b/(2a), and its y-coordinate is
When a > 0 (parabola opens upward), the vertex is the lowest point. The parabola crosses the axis only if that lowest point is at or below y = 0 — which happens exactly when -D/(4a) \leq 0, i.e., D \geq 0. When a < 0 (parabola opens downward), the vertex is the highest point, and the same logic works in reverse. Either way, D decides whether the parabola reaches the axis.
The deeper point: the discriminant is not just a formula trick. It is measuring the vertical gap between the vertex of the parabola and the x-axis. That gap is |D|/(4|a|). When the gap is zero, the vertex sits right on the axis and you get a repeated root. When the gap is positive (vertex on the correct side), the parabola crosses the axis at two points. When the gap goes the wrong way, the parabola misses the axis entirely.
The formal definition
Discriminant of a quadratic
For a quadratic equation ax^2 + bx + c = 0 with a \neq 0, the discriminant is
The nature of the roots depends on D:
| Value of D | Nature of roots | Geometrical meaning |
|---|---|---|
| D > 0 | Two distinct real roots | Parabola cuts the axis at two points |
| D = 0 | One repeated (equal) real root | Parabola touches the axis at the vertex |
| D < 0 | Two complex conjugate roots | Parabola does not meet the axis |
Additionally, when D > 0 and D is a perfect square (with a, b, c rational), the roots are rational. When D > 0 but not a perfect square, the roots are irrational.
The rational-versus-irrational distinction is worth knowing for competitive exams. If a, b, c are integers and D = 49, the roots are rational (since \sqrt{49} = 7). If D = 48, the roots involve \sqrt{48} = 4\sqrt{3}, which is irrational. The formula x = (-b \pm \sqrt{D})/(2a) makes this visible: rational roots require \sqrt{D} to be rational, which means D must be a perfect square.
A closer look at each case
Case 1: D > 0 — two distinct real roots
Take x^2 - 7x + 10 = 0. Here a = 1, b = -7, c = 10.
Since D = 9 > 0, there are two distinct real roots. And since 9 = 3^2 is a perfect square, the roots are rational:
The distance between the roots is \sqrt{D}/|a| = 3/1 = 3 — and indeed, 5 - 2 = 3. The discriminant literally encodes the gap between the two roots.
Case 2: D = 0 — one repeated root
Take x^2 - 6x + 9 = 0. Here a = 1, b = -6, c = 9.
The formula gives x = 6/2 = 3. Both roots are 3. You can see this directly by factoring: x^2 - 6x + 9 = (x - 3)^2, which is zero only when x = 3.
Geometrically, the parabola y = (x - 3)^2 has its vertex at (3, 0) — sitting exactly on the x-axis.
Case 3: D < 0 — complex roots
Take x^2 + 2x + 5 = 0. Here a = 1, b = 2, c = 5.
Negative discriminant. The formula gives
The two roots are -1 + 2i and -1 - 2i — a complex conjugate pair. They share the same real part (-1) and have opposite imaginary parts (\pm 2). No real number satisfies the equation.
An interactive discriminant explorer
Drag the point below to change the value of c in x^2 - 4x + c = 0. Watch the discriminant D = 16 - 4c change sign as c crosses 4, and see the roots merge and then vanish.
Two worked examples
Example 1: Determine the nature of roots of 3x² − 2x − 5 = 0
Step 1. Identify a, b, c.
Why: compare with ax^2 + bx + c = 0. The sign on c is negative — this will make -4ac positive, pushing D higher.
Step 2. Compute the discriminant.
Why: D = 64 > 0, so two distinct real roots. And 64 = 8^2, a perfect square — so the roots are rational.
Step 3. Find the roots.
Step 4. Separate.
Result. Two distinct rational roots: x = \dfrac{5}{3} and x = -1.
The discriminant told you everything before you even applied the formula: two real roots, both rational. That is the power of checking D first.
Example 2: For what values of k does kx² + 4x + 1 = 0 have equal roots?
This is a different kind of question — not "solve the equation" but "find the condition on a parameter." The discriminant is the perfect tool.
Step 1. Identify a, b, c in terms of k.
Why: the equation has k as the leading coefficient. For the equation to be quadratic at all, you need k \neq 0.
Step 2. Write the discriminant.
Why: equal roots happen exactly when D = 0. That is the condition you need to solve.
Step 3. Set D = 0 and solve.
Step 4. Verify. With k = 4, the equation is 4x^2 + 4x + 1 = 0, which factors as (2x + 1)^2 = 0, giving x = -1/2 as a repeated root. Correct.
Result. The equation has equal roots when k = 4.
Problems like this — "find the value of a parameter that forces a specific nature of roots" — appear constantly in board exams and JEE. The method is always the same: write D in terms of the parameter, then impose the condition D > 0, D = 0, or D < 0 as needed.
The distance between roots
There is a clean relationship between D and the gap between the two roots. If the roots are r_1 and r_2, then
This follows directly from the formula: subtracting the two roots cancels -b and doubles \sqrt{D}, giving 2\sqrt{D}/(2a) = \sqrt{D}/a, and taking the absolute value gives \sqrt{D}/|a|.
So a larger D means the roots are farther apart, and the parabola dips deeper below the axis (if a > 0). A smaller positive D means the roots are close together and the parabola barely touches the axis. At D = 0, the roots merge. The discriminant is a measure of how "spread out" the roots are.
Rational versus irrational roots
When a, b, c are all rational numbers (which they usually are in school problems), there is a finer classification:
- D > 0 and D is a perfect square: the roots are rational. Example: D = 25, roots involve \sqrt{25} = 5.
- D > 0 and D is not a perfect square: the roots are irrational, and they come in conjugate pairs p + q\sqrt{D} and p - q\sqrt{D}. Example: D = 12, roots involve \sqrt{12} = 2\sqrt{3}.
The irrational roots always come in pairs — if p + \sqrt{D}/(2a) is a root, so is p - \sqrt{D}/(2a). You cannot have one rational root and one irrational root when a, b, c are rational. This is because the \pm in the formula forces the two roots to be symmetric around -b/(2a).
Similarly, complex roots always come in conjugate pairs: if \alpha + \beta i is a root, then \alpha - \beta i is also a root. This is a deep fact — it holds not just for quadratics but for all polynomials with real coefficients.
Common confusions
-
"D = 0 means no roots." The opposite: D = 0 means there is a root — exactly one, and it is real. It is D < 0 that means no real roots.
-
"I can tell the nature of roots from the sign of c." No. A negative c does guarantee D > 0 when a > 0 (since -4ac becomes positive, adding to b^2), but a positive c tells you nothing on its own — the roots could be real, repeated, or complex depending on b.
-
"A repeated root means the equation has only one solution." Technically it has two solutions — they just happen to be the same number. This matters when you are counting roots with multiplicity (as in the fundamental theorem of algebra). The polynomial x^2 - 6x + 9 = (x - 3)^2 has x = 3 as a root of multiplicity two.
-
"D negative means the equation is wrong." The equation is fine. It simply does not have real solutions. In the complex number system, every quadratic has exactly two roots (counting multiplicity). The discriminant being negative just means those roots have non-zero imaginary parts.
-
"If D is large, the roots are large." Not necessarily. D tells you how far apart the roots are (via \sqrt{D}/|a|), not how large they are. The equation x^2 - 100x + 1 = 0 has D = 9996, which is large, but one of its roots is very close to zero (x \approx 0.01).
Going deeper
If you came here to learn what the discriminant is, how to compute it, and what it tells you about the roots, you have all of that — you can stop here. The rest is for readers who want to see the discriminant from a more structural perspective.
The discriminant as a polynomial in the coefficients
The discriminant D = b^2 - 4ac is itself a polynomial — in the coefficients a, b, c, not in x. It is quadratic in b, linear in a, and linear in c. This means that if you hold a and c fixed and vary b, the discriminant traces a parabola in the (b, D) plane, opening upward. The roots of that parabola are b = \pm 2\sqrt{ac} (when ac > 0), which are the values of b that make D = 0.
Discriminants for higher-degree polynomials
The idea of a discriminant generalises beyond quadratics. A cubic ax^3 + bx^2 + cx + d = 0 has a discriminant too:
It is messier, but it plays the same role: \Delta > 0 means three distinct real roots, \Delta = 0 means at least two roots coincide, and \Delta < 0 means one real root and a conjugate pair of complex roots. The discriminant is always a polynomial in the coefficients, and it always detects repeated roots — in fact, \Delta = 0 exactly when the polynomial has a repeated root. That is the unifying principle across all degrees.
Connection to the vertex form
You can rewrite any quadratic ax^2 + bx + c in vertex form:
The term -D/(4a) is the y-coordinate of the vertex. This makes the discriminant's role transparent: D/(4a) is how far the vertex sits from the x-axis. Positive D (with a > 0) means the vertex is below the axis; zero D means it is on the axis; negative D means it is above. The vertex form is the quadratic with the discriminant exposed.
Where this leads next
- Quadratic Formula — the formula itself, fully derived, with detailed worked examples.
- Sum and Product of Roots — Vieta's formulas let you read the sum and product of the roots directly from the coefficients, without solving.
- Quadratic Expression and Function — vertex form, axis of symmetry, and the parabola as a function rather than an equation.
- Complex Numbers — Introduction — where the roots go when the discriminant is negative.
- Quadratic Equations — Introduction — the foundational article on what quadratic equations are and how to solve them.