The Limit Comparison Test is most useful when your series looks like a simpler, well-known series but has extra terms that make a direct, term-by-term comparison awkward. If you can look at a series and say “this behaves like 1/n² as n gets large,” but can’t easily prove that every term is less than or greater than 1/n², the Limit Comparison Test lets you formalize that intuition with a single limit calculation.
How the Test Works
You start with two series whose terms are all positive: the series you’re investigating (call its terms aₙ) and a comparison series you already know converges or diverges (call its terms bₙ). Compute the limit of aₙ/bₙ as n approaches infinity. Three outcomes are possible:
- The limit is a positive, finite number. Both series do the same thing. If one converges, so does the other. If one diverges, so does the other.
- The limit is zero. The comparison series (bₙ) is growing faster, so if bₙ converges, aₙ converges too. You can’t conclude anything if bₙ diverges.
- The limit is infinity. Your series (aₙ) is growing faster, so if bₙ diverges, aₙ diverges too. You can’t conclude anything if bₙ converges.
The positive, finite case is the one you’ll use most often. It’s the cleanest: the two series are essentially the same size for large n, so they share the same fate.
When It’s the Best Choice
The Limit Comparison Test shines in a few specific situations that come up repeatedly in calculus courses.
Rational Expressions in n
If your series has terms like (3n² + 5) / (7n⁴ + 2n + 1), the dominant behavior for large n is determined by the highest powers of n in the numerator and denominator. Here that’s n² / n⁴ = 1/n². But actually proving that every term is smaller than some constant times 1/n² (which the Direct Comparison Test requires) means wrestling with inequalities. The Limit Comparison Test skips that entirely. You just compute the limit of aₙ / (1/n²), get a finite positive number, and you’re done.
Expressions With Roots or Radicals
Series with terms like 1 / √(n³ + 4n) behave like 1 / n^(3/2) for large n, but the extra “+4n” under the radical makes a direct inequality comparison messy. Taking the limit of the ratio cleans this up in a few lines of algebra.
When the Direct Comparison Inequality Goes the Wrong Way
This is the most common reason students reach for the Limit Comparison Test. Suppose you suspect a series converges, so you want to show its terms are smaller than those of a known convergent series. But when you try, the inequality actually goes the other direction: your terms are slightly larger. That doesn’t help with the Direct Comparison Test. The Limit Comparison Test doesn’t care about which series is bigger term by term. It only cares about the ratio’s long-run behavior, so the direction of the inequality is irrelevant.
When Not to Use It
The test requires all terms to be positive. If your series has terms that alternate in sign, you can’t apply it directly. One workaround: if you can show the series of absolute values converges using the Limit Comparison Test, then the original series converges absolutely, which guarantees convergence.
It’s also not the right tool when there’s no obvious comparison series. If your terms involve factorials (like n! in the numerator or denominator) or exponential functions (like 2ⁿ), the Ratio Test is almost always easier. The Limit Comparison Test works best when the dominant behavior is a power of n, because p-series (1/nᵖ) give you a convenient, well-understood family to compare against.
And if you can easily set up a direct inequality, the Direct Comparison Test is faster since it doesn’t require computing a limit at all. The Limit Comparison Test is the backup for when that inequality is hard to establish, not a replacement for every comparison.
How to Pick the Right Comparison Series
The key skill is identifying the dominant terms. For any expression in n, ask yourself: as n gets huge, which pieces actually matter?
In a fraction, keep only the highest power of n in the numerator and the highest power in the denominator. Everything else becomes negligible. So (5n³ + 12n) / (2n⁵ + n² + 7) simplifies to 5n³ / 2n⁵ = 5/(2n²), and your comparison series is 1/n². The constant 5/2 doesn’t affect convergence.
Under a square root, the same principle applies. √(9n⁴ + n) behaves like √(9n⁴) = 3n² for large n. If your term is, say, n / √(9n⁴ + n), it behaves like n / 3n² = 1/(3n), and you’d compare against the harmonic series 1/n (which diverges).
Once you’ve identified the comparison series, the limit calculation is usually straightforward. Divide the original term by the comparison term, simplify, and take the limit. If you get a positive finite number, you have your answer. If you get zero or infinity, check whether the conclusion still applies based on the rules above, or reconsider your choice of comparison series.
A Quick Decision Checklist
When you’re staring at a series on an exam, here’s a practical way to decide if the Limit Comparison Test is your move:
- All terms positive? If not, consider absolute convergence first or use the Alternating Series Test.
- Terms look like a p-series (1/nᵖ) for large n? If yes, the Limit Comparison Test is probably the fastest path.
- Direct inequality hard to set up? This is the classic signal to switch from the Direct Comparison Test to the Limit Comparison Test.
- Factorials or exponentials involved? Use the Ratio Test instead.
The Zero and Infinity Cases in Practice
Most textbook problems land in the “positive finite limit” case, but the zero and infinity cases do appear and are worth understanding intuitively. When the limit of aₙ/bₙ is zero, your series’ terms are shrinking much faster than the comparison series. So if even the slower-shrinking comparison series converges, yours certainly does too. But if the comparison series diverges, that tells you nothing, because your series might still shrink fast enough to converge.
The infinity case is the mirror image. Your terms are much larger than the comparison terms. If the comparison series diverges, yours is even worse and also diverges. But if the comparison series converges, you can’t conclude anything, because your larger terms might still be too big.
In practice, if you hit zero or infinity and can’t draw a conclusion, it usually means you picked the wrong comparison series. Go back, re-examine the dominant terms, and try again with a closer match.