How to Find the Range of a Function From a Graph: Read the Vertical Spread in Four Checks

The range is the set of all y-values the graph actually reaches. To find it from a graph, scan the picture from bottom to top, note the lowest and highest y-values the curve touches, then adjust for open circles, gaps, and asymptotes. Two functions can look similar and still have different ranges. A parabola opening upward from its vertex at y = −3 has a range of [−3, ∞), while the same parabola's mirror image opening downward has a range of (−∞, −3]. The shape alone tells you nothing. The direction and the endpoints do.

This guide covers the reading method, the notation that trips people up, and the edge cases — holes, split graphs, and horizontal asymptotes — that a quick visual scan misses.

Scan Bottom to Top to Find the Lowest and Highest Y-Values the Curve Reaches

Every range question reduces to one physical action: find the bottom of the graph and the top of the graph, then ask whether each end is included.

Step 1 — Identify the vertical axis and its scale. The y-axis is vertical. Read the tick marks first. If each gridline represents 2 units, a curve that appears to stop at the third line is at y = 6, not y = 3. Misreading the scale is the most common error on graph-reading problems in algebra and precalculus courses.

Step 2 — Trace the curve downward until it stops or turns. Note the y-coordinate at that lowest point. A parabola in vertex form f(x) = a(x − h)² + k has its minimum (if a > 0) or maximum (if a < 0) exactly at y = k, so reading the vertex off the graph gives the endpoint directly.

Step 3 — Trace upward and note the highest y-value. For a line with nonzero slope, there is no highest point — the line keeps rising or falling forever in one direction.

Step 4 — Check whether each endpoint is a solid dot or an open circle. A solid dot means the value is in the range. An open circle means it is not, and the interval uses a parenthesis on that side.

Step 5 — Write the interval from lowest y to highest y. Use square brackets [ ] for included endpoints and parentheses ( ) for excluded ones. The infinity symbol ∞ always takes a parenthesis because infinity is a direction, not a number you can touch.

If the lowest point is y = −2 with a solid dot and the graph rises forever, the range is [−2, ∞). If that same point were an open circle, it becomes (−2, ∞). One pencil mark changes the answer.

A failure case worth knowing before an exam: a graph with a solid dot at the bottom and an open circle at the top but no vertical gridlines labeled. Students often assume the gridlines count by 1. Confirm the axis scale before writing any interval.

Determine Whether the Endpoints Are Included With Bracket and Parenthesis Rules

Notation is where most points get lost on tests, even when the visual reading is correct. Interval notation has a fixed grammar and three common mistakes.

The first rule to internalize: brackets always face the number they include. A bracket that faces away from the number, like ]3, means 3 is excluded — that's a typo, not valid notation. Every interval must be written with the smaller value on the left.

The second rule: ∞ and −∞ always take parentheses. Writing [3, ∞] is a notation error that some graders mark wrong even when the concept is right.

The third rule: never bracket an asymptote. If a curve approaches y = 2 but never reaches it, the range excludes 2. That is a parenthesis, not a bracket. Many textbooks distinguish between "the graph appears to touch the line" and "the graph reaches the line" — on a hand-drawn graph, assume it approaches unless a solid dot is clearly shown at the intersection.

Read Parabolas in Both Directions by Locating the Vertex

A parabola has exactly one turning point, which makes it the cleanest graph to read — the vertex is both the visual peak or valley and the y-value that anchors the interval.

For f(x) = x² − 4x + 1, completing the square gives f(x) = (x − 2)² − 3. The vertex is (2, −3) and the parabola opens upward because the coefficient of x² is positive. So the lowest y-value is −3, it's included, and the range is [−3, ∞).

Flip the sign: g(x) = −(x − 2)² − 3 opens downward with vertex at (2, −3). Now the highest y-value is −3, and the range is (−∞, −3]. Same vertex, opposite range. The coefficient on the squared term controls the direction, and the direction controls which side of the vertex is unbounded.

Worked example. A graph shows a parabola with vertex at (1, 2) opening downward and crossing the x-axis at x = −1 and x = 3. Find the range. Solution: the vertex is the highest point at y = 2, it's a solid point on the curve (the vertex of a parabola is always part of the graph), and the arms extend downward forever. Range: (−∞, 2].

The vertex always belongs to the graph of a parabola — that's a structural fact, not a reading judgment. So the bracket at the vertex side is always square, no exceptions.

Spot Asymptotes That Leave Gaps in the Range Even When the Curve Looks Continuous

Rational functions produce the trickiest range problems because the graph can look like it covers every y-value while actually skipping one or more.

Take f(x) = 1/x. As x grows large in either direction, y gets close to 0 but never equals it. As x approaches 0 from either side, y shoots toward ∞ or −∞. The graph has a horizontal asymptote at y = 0 and a vertical asymptote at x = 0. The range is (−∞, 0) ∪ (0, ∞). Notice the union symbol — the range is two separate pieces because 0 is missing.

When does a horizontal asymptote affect the range? Usually when the numerator's degree is less than or equal to the denominator's. For f(x) = (2x + 1)/(x − 3), the degrees match, so the horizontal asymptote is y = 2 (the ratio of leading coefficients: 2/1). The range is all real numbers except 2, written (−∞, 2) ∪ (2, ∞).

Linear-over-linear rational functions whose numerator has degree 1 and denominator has degree 1 (with the denominator not dividing the numerator evenly) always have a horizontal asymptote at some y = c, and the range excludes c. This is one general fact worth memorizing because it appears on nearly every precalculus exam.

The range of a rational function can also have vertical gaps in y-space, not just horizontal ones — but those only happen when the horizontal asymptote equals a value the function still reaches elsewhere. Check carefully on a printed graph whether the curve actually crosses the asymptote line. Some rational functions cross a horizontal asymptote and later approach it from the other side, in which case the asymptote value is in the range.

Recognize the Four Core Function Families Whose Ranges You Can Predict From Shape

Four families cover most high school and early college graph-reading problems. Once you can recognize the shape, the range follows almost automatically.

A linear function f(x) = mx + b whose slope m ≠ 0 has range all real numbers, (−∞, ∞). A horizontal line f(x) = c has range {c}, a single value — a special case that confuses people who expect a range to be an interval.

A quadratic function has range [k, ∞) when it opens upward and (−∞, k] when it opens downward, where k is the vertex's y-coordinate.

An absolute value function f(x) = |x − h| + k has range [k, ∞). It looks like a V and its vertex behaves exactly like a parabola's vertex in the range calculation.

An exponential function f(x) = a · bˣ (with a > 0 and b > 0, b ≠ 1) has range (0, ∞). It has a horizontal asymptote at y = 0 and climbs forever in one direction. The exponential's range is often mistakenly written as [0, ∞) because students see the curve leveling off near y = 0 — but it never reaches 0, so the lower bound is always a parenthesis.

If the graph you're reading matches one of these basic shapes but is shifted, stretched, or flipped, the range shifts and stretches with it. The vertex or asymptote is the anchor.

Apply the Vertical Line Test in Reverse to Confirm the Graph Is a Function Before Reading Its Range

You can't ask for a function's range if the graph isn't a function. Before you compute anything, check whether every x-value has at most one y-value.

Step 1 — Slide a vertical line across the graph from left to right. If at any x-position the line touches the graph at more than one point, the graph is not a function.

Step 2 — If the graph fails the test, don't ask for a range. The term "range" only applies to functions. For a general relation, the y-values present are called the range of the relation, but most curricula stop using the term once the vertical line test fails. If the problem asks for the range of a circle, for example, you're being asked about the set of y-values it occupies, not the range of a function.

Step 3 — If the graph passes, apply the bottom-to-top scan. The vertical line test does not rule out gaps, branches, or asymmetry — it only confirms single-valuedness.

Step 4 — Check whether a restricted domain has rescoped the graph. A circle cut in half, a parabola restricted to x ≥ 0, or an exponential truncated at x = 2 changes the range along with the visible portion. Only the visible portion counts.

Failing the test happens most often with circles, sideways parabolas, and ellipses. These do have a range as a relation (the set of y-values in the picture), but the four end-of-chapter tests usually want a function's range, so the vertical line test is the gate.

Combine Two Intervals With Union Notation When the Graph Splits

Piecewise functions and absolute values rewritten as piecewise pieces can produce a range made of two or more separate intervals. Write them with the union symbol ∪ and list them from left to right in value order.

Take f(x) = |x| + 1. Its graphic is a V with vertex at (0, 1). Range: [1, ∞). That's a single interval. Now take g(x) = 1 − |x|. That's an upside-down V with vertex at (0, 1). Range: (−∞, 1]. Still a single interval.

Split ranges happen with piecewise definitions like f(x) = x for x ≤ 0 and f(x) = x + 3 for x > 0. On the left piece, range is (−∞, 0]. On the right piece, range is (3, ∞). Combined, the range is (−∞, 0] ∪ (3, ∞). The gap between 0 and 3 (exclusive) is not covered.

When writing the union, start with the lower-valued interval. Common mistake: writing "0 to 3 and (−∞, ∞)" — that's a false range because it double-counts. Only interval pieces with no y-value overlap should be joined by ∪. If two pieces overlap in y-values, merge them into a single interval first.

A hole in a graph (a removed point) works the same way. If the graph is continuous except at one point, remove that y-value from the interval. For a line f(x) = 2x − 1 with a hole at x = 3, the y-value 5 is missing, so the range is (−∞, 5) ∪ (5, ∞).

Whether a graph has an open circle, a hole, or an asymptote, the same visual tool works: locate the y-value that's missing, and break the interval there.

Estimate Carefully From Graphs That Lack Gridlines or Use Unequal Scales

Real test graphs sometimes skip gridlines or stretch one axis. Reading the range correctly depends on reading the scale correctly.

If a graph is printed with no y-axis numbered, compare the bottom of the graph to the x-axis line. Is the lowest point at the axis, below it, or above it? If there's no axis drawn, the problem almost always states a reference somewhere — check the function rule if one is given. A hand-drawn sketch without any scale is not a graph you can reliably read.

A subtler trap: unequal spacing. Some histograms, log plots, and function plots change spacing on the y-axis. Check whether each gridline represents the same vertical distance. If not, the "visual bottom" is not the "numeric bottom."

Another trap: estimation rounding. If a curve appears to reach the y = 3 line but is actually reaching y = 2.98, writing [3, ∞) is off by 0.02. Use the tick marks — if the lowest point is at or near a labeled line, bracket at that labeled value. If it's between lines, choose the closest labeled value or use the function rule to calculate exactly.

Address These Six Range-Mistake Patterns Before You Submit an Answer

The following mistakes account for most lost points on range problems involving graphs. Each has a spoke-specific fix.

Bracket at infinity. Writing [2, ∞] instead of [2, ∞). Infinity is a direction, not a number. Fix: change the closing bracket to a round parenthesis.

Reading the x-range instead of the y-range. The interval on the x-axis is the domain, not the range. Fix: read vertically, not horizontally.

Assuming y = 0 is in the range. Many graphs approach y = 0 but never touch it, especially exponentials and rationals. Fix: check for a horizontal asymptote or dashed line at the would-be boundary.

Ignoring open circles. An open circle at the bottom of a parabola means the vertex is excluded, so the range becomes (k, ∞). Fix: look at the dot's fill, not just its position.

Missing union intervals. When the graph has two disconnected pieces, one interval is incomplete. Fix: trace the whole graph from bottom to top, check for gaps, and union the pieces.

Confusing the range with the codomain. The codomain is the set a function is declared to map into (usually all real numbers), not what the graph actually reaches. Fix: report only what the picture shows.

Keep a short checklist taped to your notebook for the first several problems. The patterns fade once you've practiced them — but they don't fade without practice.

FAQ

Does the vertex of a parabola always belong to the range?

Yes for any standard parabola on a graph. The vertex is the graph's own point because every x-value in the domain yields a y-value, and the vertex's x-coordinate is inside the parabola's domain — all real numbers. The bracket at the vertex side is always square. What can fail is an endpoint elsewhere on a restricted-domain problem.

Can a linear function have a range like [3, ∞)?

Not if the slope is nonzero and the domain is all real numbers. A nonhorizontal line continues in both y-directions forever, so its range is (−∞, ∞). A [3, ∞) range on a linear-looking graph means either the domain has been restricted or you're accidentally reading a ray, not a full line.

Does a horizontal asymptote always remove a value from the range?

No — only if the curve never reaches it. The graph of f(x) = (x² − 1)/(x² − 1) is the line y = 1 with a hole at x = ±1, and the range is {1}, asymptote included. Check for an actual crossing before excluding an asymptote value. The default assumption on standardized problems is that horizontal asymptotes are not crossed, but verifiable graphs exist where they are.

Why is the range of an exponential function written with a parenthesis at 0?

Because the exponential curve has y = 0 as a horizontal asymptote that it never touches. At any finite x-value, bˣ is positive — no power of a positive base is zero. The range is (0, ∞), not [0, ∞). This matters on standardized exams where "[0, ∞)" is a common distractor.

If a graph's range is (−2, 3), what does it look like?

The graph's lowest y-value is a bit above −2 (approached but not reached, likely from an open circle or asymptote), and its highest y-value is a bit below 3 (also excluded). The graph sits entirely between those two horizontal lines without touching either one. Bracket endpoints are the difference between (−2, 3) and [−2, 3].

How do you find the range when the graph has two separate curves?

Identify each curve's own range as if it were alone, then union them. If the two ranges overlap, merge them into one interval. If they're separated by a gap (like (−∞, 0] and (3, ∞)), join the two with the union symbol. Never pick one curve and call that the answer — the graph's range is all its y-values.

What's the range of a horizontal line y = 5?

The range is a single point: {5}, written in set braces. A horizontal line's y-value never changes, so there's exactly one output. The set-brace notation distinguishes this case from the interval [5, 5], which is also correct but less common — both express that 5 is the only y-value.

Does the range depend on how much of the graph is drawn?

No — the range is a property of the function on its full domain, not of the picture. A parabola drawn only from x = −2 to x = 2 still has range [0, ∞) if its vertex is at (0, 0) and its algebraic form has no domain restriction. What you see is a window; the range describes the whole object. On a restricted-domain problem, though, only the displayed portion counts, and the range shrinks with the domain.