IRR vs. NPV: Which Should You Use to Evaluate an Investment?
September 22, 2026
NPV gives a dollar amount showing how much value an investment is expected to add. IRR gives a percentage — the rate at which NPV would equal exactly zero. They usually agree on whether to accept a single investment, but when comparing two different-sized investments, they can rank them in opposite orders — and when they disagree, NPV is the more reliable metric.
Skip the math and get your answer instantly:
Open the IRR Calculator →A Case Where IRR and NPV Disagree
Project A: invest $10,000, receive $15,000 in one year. Project B: invest $100,000, receive $130,000 in one year. Both discounted at 10%.
| Project | IRR | NPV (at 10%) |
|---|---|---|
| A | 50% | +$3,636 |
| B | 30% | +$18,182 |
Which One Should You Actually Pick?
By IRR alone, Project A looks far better (50% vs. 30%). But Project B creates more than five times as much actual dollar value ($18,182 vs. $3,636), because IRR is a percentage that ignores scale entirely. If you have $100,000 to invest and can only choose one project, Project B is the better choice despite its lower IRR — this is exactly the kind of conflict where NPV, not IRR, should guide the decision.
When IRR and NPV Usually Agree
For a single, standalone investment with conventional cash flows (one upfront cost, then positive returns), IRR and NPV always agree on the basic accept/reject decision: if IRR is above your required rate of return, NPV at that same rate will be positive, and vice versa. Conflicts only arise when ranking or choosing between multiple, differently-scaled projects.
The Multiple-IRR Problem
If a project's cash flows change sign more than once — an initial cost, followed by profits, followed by another large cost later (common in projects with cleanup or decommissioning costs) — the math can produce more than one rate where NPV equals zero, giving multiple 'valid' IRRs with no clear answer as to which one is meaningful. NPV doesn't have this problem, since it's a single well-defined number at any given discount rate.
Frequently Asked Questions
Why did the project with the lower IRR turn out to be the better choice?
Because IRR is a percentage that says nothing about the scale of the investment. A small project can have an impressively high IRR while creating relatively little total value, while a larger project with a more modest IRR can create far more value in dollar terms — which is what NPV actually measures.
When do IRR and NPV give the same accept/reject answer?
For a single standalone investment with a conventional cash flow pattern (one upfront cost followed by positive returns), IRR above your required rate always corresponds to a positive NPV at that same rate — they only diverge when ranking or choosing between multiple different-sized projects.
Why does a project sometimes have more than one IRR?
When cash flows change sign more than once (cost, then profit, then another cost), the NPV-vs-rate relationship can cross zero multiple times, producing more than one mathematically valid IRR — NPV avoids this ambiguity since it's a single value at any chosen discount rate.
Which metric do most finance professionals prefer?
NPV is generally considered the more theoretically sound metric for investment decisions, since it directly measures value creation in dollar terms — IRR remains popular because a percentage is intuitive to communicate, but it's typically used alongside NPV, not as a replacement for it.
Can I use IRR to compare investments with different time horizons?
IRR can be especially misleading here — a short project with a high IRR may look better than a longer project with a lower IRR but much greater total value creation. Comparing NPVs at a consistent discount rate is usually the more reliable approach across different time horizons.