Finding the Vertical Asymptote of a Function

Start With the Simplified Denominator

To find vertical asymptotes, work with a rational function, one polynomial divided by another, and apply a single rule: simplify first, then set the remaining denominator equal to zero. The vertical asymptote is the vertical line x = a at those zeros. The function is undefined at x = a, and its values grow without bound as x gets closer to a from either side. You are looking for the real numbers that make the simplified denominator zero, because those are the only places a rational function can blow up. If the denominator has no real zeros after simplification, there is no vertical asymptote. If a factor cancels completely, that x-value is a hole, not a vertical asymptote.

Factor, Cancel, Then Test

The vertical asymptote rules are precise, and they all flow from one instruction: factor both numerator and denominator, cancel common factors, then check what remains. After simplification, every distinct real root of the denominator is a vertical asymptote, provided it is not also a root of the numerator. If numerator and denominator share a factor, that shared root is a hole, not an asymptote. So x = 1 is a hole, not a vertical asymptote. The rule fails if you skip the cancellation: setting the original denominator to zero would give you x = 1, but that is wrong. You must simplify before you apply the zero-denominator test.

Work through three examples to see the rule in action. First, a linear denominator: the denominator is already factored, and nothing cancels. Both are vertical asymptotes, one per distinct real root. Second, a quadratic denominator that factors into two distinct linear factors: each real root becomes a vertical asymptote. Third, a quadratic denominator with no real roots: there is no vertical asymptote, even though the denominator is zero nowhere on the real line. The pattern is consistent: simplify, find real roots of the remaining denominator, and those roots are your vertical asymptotes.

Once you know a vertical asymptote exists at x = a, you need to know which way the graph goes on each side. The one-sided behavior determines whether the function approaches +∞ or -∞ as x approaches a from the left and from the right. The two sides can go to the same infinity or opposite infinities. The sign chart is the tool that settles it, and you build it from the simplified function, not the original, because the canceled factor would mislead you.

The vertical asymptote limit definition is what makes the rule rigorous: a vertical asymptote at x = a means at least one one-sided limit of f(x) as x approaches a is ±∞. In symbols, lim f(x) = +∞ as x→a⁺ or lim f(x) = -∞ as x→a⁻, and similarly for the other side. This is an infinite limit, not a finite one. The function values increase or decrease without bound, which is different from a finite limit like y = 3. The limit does not exist in the usual sense because it is not a real number. For a rational function, this infinite limit occurs exactly at the zeros of the simplified denominator. The formal definition also explains why a graph can never cross a vertical asymptote: the function is undefined at x = a, so there is no point on the graph with that x-coordinate. The line x = a is not part of the graph; it is a boundary the curve approaches without touching.

A graph cannot cross a vertical asymptote because the function is undefined there. At x = a, the denominator is zero after simplification, so f(a) does not exist. There is no point (a, y) on the curve, which means the graph has no value to cross the line with. This is different from a horizontal asymptote, where the graph can cross the line at finite x-values because the horizontal asymptote only constrains end behavior as x goes to positive or negative infinity. For a vertical asymptote, the line x = a is a barrier in the domain itself: the function is not defined there, so the curve is split into two separate branches, one on each side of the line. Each branch approaches the line vertically, going up or down without bound. The graph never touches x = a because touching would require a defined point, and there is none. This is why the vertical asymptote is a hard boundary, unlike the soft approach of a horizontal one.

When a factor cancels completely, the result is a hole, a single point missing from the graph, not a vertical asymptote. If the factor does not cancel, that x-value is a vertical asymptote. The rule is: every real root of the original denominator is excluded from the domain, but only those that survive simplification become vertical asymptotes. First, failing to factor before canceling: you cannot see what cancels if you do not factor completely. Second, setting the original denominator to zero without simplifying: this turns a hole into a false vertical asymptote. Third, assuming both sides of a vertical asymptote go to the same infinity: they can go to +∞ on one side and -∞ on the other, as with g(x) = 1/(x - 5).Fifth, confusing the y-value of a hole: you must use the simplified function, not the original, because the original is undefined there. If you are working a problem and get a vertical asymptote at a value where the numerator is also zero, go back and factor again, you likely missed a cancellation. The sign chart fixes the direction, and the simplification fixes the location.

When the denominator does not factor nicely, you still apply the same rule, but you may need the quadratic formula. For a quadratic denominator like 2x² + 3x - 5, factor it if you can; if not, use the quadratic formula to find the roots. Each real root is a candidate vertical asymptote, provided the numerator does not share that root. If the discriminant is negative, there are no real roots, and there is no vertical asymptote. The vertical asymptote rules do not depend on the complexity of the denominator, only on whether it has real zeros after simplification.

Vertical asymptotes are only one part of a rational function's graph. If the numerator degree is exactly one more, there is a slant asymptote, which is a linear function the graph approaches at the tails. If the degree difference is two or more, there is no linear asymptote. These horizontal and slant asymptotes can be crossed at finite x-values, unlike vertical asymptotes. The vertical asymptote rules as stated apply to rational functions, where both numerator and denominator are polynomials. Other function types can have vertical asymptotes too, but the rules differ.Exponential and logarithmic functions have horizontal or vertical asymptotes under specific conditions: log(x) has a vertical asymptote at x = 0, and e^x has a horizontal asymptote at y = 0. These are not rational functions, so the factor-and-cancel procedure does not apply. OpenStax Precalculus 2e, section 5.6, notes that non-rational functions may have asymptotes but the rules differ, so check the function type before applying any rule blindly.

When the normal route of factoring and canceling fails, say the denominator is irreducible over the reals, or you have a graphing problem where the algebra is messy, use a graphing calculator or software to plot the function and look for vertical lines where the graph shoots up or down. Then verify algebraically: plug in values just left and right of the suspected asymptote to see if the function values grow without bound. If the denominator has complex roots only, there is no vertical asymptote, and the graph is continuous everywhere. If a root makes both numerator and denominator zero, factor and cancel to check for a hole. This backup plan takes longer but never fails, because the definition of a vertical asymptote is about the limit being infinite, and you can always test that numerically.

After you compute a vertical asymptote, verify it by checking the limit. Pick a value just left of a and one just right, plug them into the simplified function, and compute the magnitude. This two-sided check is essential because a vertical asymptote requires at least one one-sided limit to be infinite. If both one-sided limits are finite, it is not a vertical asymptote. For example, in f(x) = (x² - 1)/(x - 1), at x = 1, the left and right limits both approach 2, so it is a hole, not a vertical asymptote. The verification step catches the most common error: treating a hole as an asymptote because you skipped the cancellation.