Excel Practice Test

โ–ถ

How Do I Round A Number In Excel โ€” Fast Answer

The short version. Use =ROUND(A1, 2) when you want a clean two-decimal currency value, =ROUNDUP(A1, 0) when you must never round down (think shipping weights and ceiling math), =ROUNDDOWN(A1, 0) when you must never inflate (tax-deductible amounts, conservative accounting). Need a number rounded to the nearest 5, 25, or quarter hour? That's =MROUND(A1, 5). Trying to chop decimals without any rounding logic at all? =INT() or =TRUNC(). The whole game is picking the right function, then feeding it a num_digits argument that tells Excel where the rounding cut-off sits. Get that part right and everything downstream just works.

ROUND: The Function Most People Actually Want

Here's the thing about Excel rounding. Most users โ€” and that's probably 80% of them โ€” only ever need ROUND. They just don't know it yet. They format cells to show two decimals, watch the display change, and assume the value changed too. It didn't. Format is cosmetic. ROUND is surgical.

Syntax: =ROUND(number, num_digits). Two arguments, both required. The first is the value (or a cell reference). The second tells Excel how many decimal places to keep. Positive numbers cut decimals โ€” =ROUND(3.14159, 2) returns 3.14. Zero rounds to a whole integer โ€” =ROUND(3.7, 0) returns 4. And here's the part that catches people off guard: negative numbers round to the left of the decimal. =ROUND(1247, -2) returns 1200. That's rounding to the nearest hundred. Negative two means "two digits to the left." Once that clicks, you stop fighting the syntax.

ROUND uses standard arithmetic rounding โ€” anything 0.5 and above rounds up, anything below rounds down. That sounds simple. It's not always what regulators want. In the United States, GAAP accounting tolerates ROUND for most general-ledger work. European IFRS standards sometimes mandate ROUNDDOWN for tax accruals and reserves, on the conservative-bias principle. If you're rounding for a regulated filing, check the rule before picking the function. Don't just default to ROUND because it's familiar.

One quirk worth knowing. Some statisticians call standard ROUND "banker's rounding" โ€” that's not quite right. True banker's rounding (also called round-half-to-even) sends exact .5 values to the nearest even integer. So 2.5 becomes 2, 3.5 becomes 4. Excel's ROUND does not do this. It rounds .5 away from zero every time. If you genuinely need banker's rounding, you'll have to build it with MROUND tricks or VBA. The R language and Python's built-in round() both default to banker's; Excel's ROUND does not. Mismatched expectations between these tools have caused some painful reconciliation meetings.

ROUNDUP, ROUNDDOWN and MROUND Compared

๐Ÿ“‹ ROUNDUP

=ROUNDUP(number, num_digits) always rounds away from zero. =ROUNDUP(2.1, 0) returns 3. =ROUNDUP(-2.1, 0) returns -3. Notice the second one โ€” "away from zero" means negatives get more negative. Use ROUNDUP when you can't underestimate. Classic case: a shipping cost where the carrier charges per pound and a 4.1 lb package costs the same as a 5 lb package. =ROUNDUP(weight, 0) gives you the billable weight. Same logic applies to printer pages, server seats, and any rate card that bills by the unit ceiling.

๐Ÿ“‹ ROUNDDOWN

=ROUNDDOWN(number, num_digits) always rounds toward zero. =ROUNDDOWN(2.9, 0) returns 2. =ROUNDDOWN(-2.9, 0) returns -2. This is the conservative direction โ€” useful for tax deductions where overclaiming triggers an audit, or for inventory pulls where you don't want phantom stock. ROUNDDOWN truncates magnitude, not sign. Don't confuse it with INT, which floors regardless of sign (more on that in a minute).

๐Ÿ“‹ MROUND

=MROUND(number, multiple) snaps a value to the nearest multiple of whatever you specify. =MROUND(34, 5) returns 35. =MROUND(34, 10) returns 30. =MROUND(7.23, 0.25) returns 7.25. This function shines anywhere humans need clean increments. Pricing to the nearest nickel, dime, or quarter. Scheduling to the nearest 15 minutes: =MROUND(time_value, TIME(0,15,0)). Hours logged in 0.25 chunks: =MROUND(hours, 0.25). The number and the multiple must share a sign โ€” both positive or both negative โ€” or MROUND returns #NUM!. A small gotcha but easy to fix with ABS.

CEILING and FLOOR: Multiple-Based Rounding with Direction

CEILING and FLOOR are MROUND's bossier siblings. They also round to a multiple, but they force the direction. =CEILING(number, significance) always rounds up to the next multiple. =FLOOR(number, significance) always rounds down. =CEILING(23, 5) returns 25. =FLOOR(23, 5) returns 20. Where MROUND picks whichever multiple is closer, CEILING and FLOOR pick a side and commit.

These two have an awkward history with negative numbers. The legacy versions throw errors or behave unpredictably below zero. Microsoft fixed it with CEILING.MATH and FLOOR.MATH โ€” these accept negatives and let you control direction with a third mode argument. If you're working on Excel 2013 or later, default to the .MATH versions. They're safer, they're consistent, and they support the same significance logic.

Practical examples. A delivery service charges per 5-kg increment โ€” =CEILING(weight, 5) gives the billed bracket. A bank teller breaks cash withdrawals into $20 bills โ€” =FLOOR(amount, 20) tells you the largest $20-divisible payout, leaving the remainder for smaller bills or coins. A factory shipping pallets of 24 units โ€” =CEILING(order_quantity, 24) tells you how many to actually palletize. These are operational rules disguised as math.

One real-world trap. CEILING and FLOOR work on positive significance values cleanly, but negative-significance behavior differs from sheet to sheet (Excel vs Google Sheets vs LibreOffice). If your workbook gets opened in different tools, stick to positive significance and handle sign separately with IF or ABS. Saves headaches at month-end close when someone opens the model in Sheets and the totals don't reconcile.

INT vs TRUNC vs ROUNDDOWN โ€” They're Not the Same

๐Ÿ”ด INT

=INT(number). Returns the integer portion by rounding DOWN toward negative infinity. INT(2.9) = 2. INT(-2.1) = -3. That second result trips everyone. INT doesn't truncate sign-neutrally โ€” it floors. Negative numbers get pushed further from zero, not closer. Use INT only when you genuinely want mathematical floor behavior.

๐ŸŸ  TRUNC

=TRUNC(number, [num_digits]). Truncates toward zero, lopping off decimals without any rounding. TRUNC(2.9) = 2. TRUNC(-2.9) = -2. The optional second argument lets you keep some decimals: TRUNC(3.14159, 2) = 3.14. Different from ROUND because there's no rounding logic โ€” just a hard chop. Useful when you need raw integer extraction and don't want any value movement.

๐ŸŸก ROUNDDOWN

=ROUNDDOWN(number, num_digits). Identical to TRUNC for any number with a positive num_digits โ€” both lop toward zero. The difference is intent and flexibility. ROUNDDOWN accepts negative num_digits to round-down past the decimal (=ROUNDDOWN(1247, -2) = 1200). TRUNC doesn't behave well with negative num_digits. Pick ROUNDDOWN for explicit rounding intent in any direction.

EVEN and ODD: Niche but Useful

Two more rounding functions sit in the toolbox that almost nobody talks about. =EVEN(number) rounds up to the nearest even integer. =ODD(number) rounds up to the nearest odd integer. EVEN(2.1) returns 4 (not 2 โ€” it always rounds away from zero to the next even). EVEN(2) returns 2 because it's already even. ODD(2.1) returns 3.

Where do you actually use these? Packaging that requires pairs โ€” eyeglass frames, shoes, dance partners. Drug dosing that ships in even-count blister packs. Lab work that runs duplicate samples. ODD shows up in machine learning when you want guaranteed tiebreaker counts in voting ensembles (an odd number of judges can't tie). They're not glamorous, but when the requirement hits, they save a nested IF or two.

Negative behavior is consistent with the family โ€” both functions round away from zero. EVEN(-2.1) returns -4. ODD(-2.1) returns -3. Mostly intuitive, occasionally surprising. Test once with edge cases before deploying in a regulated model.

The num_digits Argument โ€” One Picture, All the Rules

This single argument controls almost every rounding function. The rule is positional, counted from the decimal point:

  • num_digits = 2 โ†’ keep 2 decimals. ROUND(3.14159, 2) = 3.14.
  • num_digits = 1 โ†’ keep 1 decimal. ROUND(3.14159, 1) = 3.1.
  • num_digits = 0 โ†’ integer. ROUND(3.7, 0) = 4.
  • num_digits = -1 โ†’ nearest ten. ROUND(1247, -1) = 1250.
  • num_digits = -2 โ†’ nearest hundred. ROUND(1247, -2) = 1200.
  • num_digits = -3 โ†’ nearest thousand. ROUND(1247, -3) = 1000.

Same logic applies to ROUNDUP, ROUNDDOWN, and TRUNC. Memorize this once and stop second-guessing it for the rest of your spreadsheet life.

Real-World Rounding Recipes That Actually Get Used

Theory's fine. Patterns are better. Here are the rounding formulas that show up over and over in production workbooks โ€” copy them, adjust the cell references, and you're done.

Currency to two decimals. =ROUND(A1, 2). The bread and butter. Every invoice, every P&L, every dashboard that displays dollars. Don't trust cell formatting to do this โ€” formatting only changes the display while keeping the full-precision value alive in the cell, and that bites you later when totals don't tie out by a penny or two. ROUND fixes the value, not just the look.

Prices ending in .99. =ROUNDDOWN(A1, 0) + 0.99. Strip the decimals, drop on the psychological-pricing tail. Used everywhere from retail catalogs to subscription SaaS pricing pages. If you want .95 endings, swap the constant. If you want .49 endings, same drill.

Hours to the nearest quarter. =MROUND(A1, 0.25). Payroll software often demands quarter-hour granularity even when timeclocks capture per-second data. MROUND with 0.25 maps 4.13 hours to 4.25 and 4.06 hours to 4.00. Done.

Times to nearest 15 minutes. =MROUND(A1, TIME(0,15,0)). Excel times are fractions of a day, so 15 minutes is TIME(0,15,0) not literally 0.25. Mix this up and you'll round to quarter-days instead of quarter-hours, which produces some entertaining timesheet bugs.

Sales to the nearest thousand for a slide. =ROUND(A1, -3). Or =ROUND(A1/1000, 0) if you're going to display "234K" โ€” round before the suffix, not after. Otherwise 1499 becomes "1K" instead of "1K" wait that's the same โ€” OK try 1500 which should round to 2K but if you do ROUND(A1, -3)/1000 you get 2 cleanly, while ROUND(A1/1000, 0) also gives 2. Both work. Pick whichever reads cleaner in your formula bar.

Tax accrual, round-down for safety. =ROUNDDOWN(A1 * tax_rate, 2). Multiply first, ROUNDDOWN second. Conservative accrual. Auditors prefer this for liability accounts because it never overstates the company's obligation.

Significance bucket for charts. =CEILING(MAX(data_range), 100). Sets the upper chart axis bound to the next clean hundred above your data peak. Same trick with FLOOR for the lower bound. Charts auto-scaled this way always look intentional. The default Excel auto-scaling sometimes picks weird axis values like 0 to 487 โ€” using CEILING gives you a clean 0-to-500 axis without manual override.

Rounding inside SUMIF or AVERAGEIF. =ROUND(AVERAGEIF(range, criteria), 2). Wrap the aggregate in ROUND when the result feeds a display cell. If the result feeds another formula, leave it unrounded โ€” premature rounding compounds errors across the calculation chain. Rule of thumb: round at the boundary between math and display, not in the middle.

Quick Pause โ€” Try a Free Excel Practice Quiz

ROUND Function vs Cell Formatting

Pros

  • ROUND changes the stored value โ€” totals tie out correctly
  • Survives copy-paste between workbooks and tools
  • Auditable: the formula bar shows what the cell actually contains
  • Works inside other formulas โ€” no precision leakage
  • Required for regulated calculations (tax, payroll, GAAP/IFRS)

Cons

  • Cell formatting only changes display โ€” value remains full precision
  • Sums of formatted cells often differ from displayed totals by a penny
  • Format gets lost when exporting to CSV or pasting as values
  • Two cells showing 1.50 might secretly hold 1.499 and 1.501
  • Hidden source of audit findings when displayed vs stored values diverge

The Gotchas That Catch Even Experienced Users

Floating-point precision is the silent killer. Excel stores numbers using the IEEE 754 double-precision standard, which means 0.1 + 0.2 doesn't equal exactly 0.3 โ€” it equals 0.30000000000000004. You never see this in normal display, but when you build a comparison like =IF(A1 = 0.3, "yes", "no") and A1 was reached by summing 0.1 three times, the test returns "no." Wrap the calculation in ROUND: =IF(ROUND(A1, 10) = 0.3, ...). Round to a few more digits than you actually care about, and the comparison stops lying.

Cell format vs cell value is the most common rounding bug in business spreadsheets. A cell shows 100.00. The actual value is 100.004. You add five of these cells together and the total shows 500.02 instead of 500.00. The display rounded each cell individually, but the underlying SUM used full precision. Auditors call this "rounding drift" and they hate it. The fix is to ROUND the source values, not the display. Or use the workbook option "Set precision as displayed" โ€” but that's destructive and global, so use it cautiously.

Negative-number rounding direction is the second-most common bug. ROUND(-1.5, 0) returns -2 (away from zero, magnitude up). INT(-1.5) returns -2 also, but for a different reason (floor toward negative infinity). TRUNC(-1.5) returns -1 (toward zero, magnitude down). ROUNDDOWN(-1.5, 0) returns -1 (toward zero). Mix these up in a sales-commission formula and someone gets paid wrong. Test with a representative negative value before shipping any rounding-sensitive model.

Date and time rounding deserves its own warning. Excel stores dates as serial numbers and times as decimal fractions of a day. =ROUND(NOW(), 0) rounds the time-fraction to zero โ€” meaning it returns today's date at midnight if it's before noon, tomorrow's date if it's after. Probably not what you wanted. For time rounding, use MROUND with TIME() values or convert to minutes first.

Locale settings can flip decimal-comma countries. In Germany, France, and most non-English Excel installations, the decimal separator is a comma, not a period. Formulas typed as =ROUND(3.14, 2) may need to become =ROUND(3,14;2). Same function, different syntax. If you're sharing workbooks across borders, test on a target machine before going live.

One more โ€” nested rounding compounds error. =ROUND(ROUND(A1, 2) * 1.075, 2) rounds twice. Mathematically that's two opportunities for tiny drift. In a single-cell model nobody notices. In a model with 50,000 line items totaling tens of millions, rounding drift can add up to real money. Round once, at the display boundary. Not at every intermediate step. This is one of those rules that feels pedantic until it's the difference between matching the bank statement and not.

Five Stages of Rounding Mastery in Excel

1

Start here. =ROUND(A1, 2) for money, =ROUND(A1, 0) for whole numbers. Cover 80% of cases and stop fighting cell formatting.

2

Learn when 'always up' or 'always down' matters. Shipping weights, tax-conservative numbers, billing units. Direction-aware rounding.

3

Quarter hours, nickel pricing, 5-degree increments. MROUND replaces a whole class of ugly nested IFs.

4

Significance-based rounding with sign control. Bucketing, bracketing, chart-axis rounding. Use the .MATH variants โ€” they handle negatives.

5

Truncation vs flooring, parity rounding, the floating-point traps. Now you can debug anyone else's rounding-related #VALUE error in under a minute.

Compliance and Accounting Standards Around Rounding

Most users never think about rounding rules from a regulatory angle, but they exist and they matter when the numbers go to auditors. Under United States Generally Accepted Accounting Principles (GAAP), standard arithmetic rounding (ROUND-style) is permitted for most operational accounting โ€” revenue recognition, expense classification, depreciation schedules. International Financial Reporting Standards (IFRS) are similar but lean conservative for tax accruals and provisions, where rounding down is the safer convention.

The European Union's VAT directives are extremely specific about rounding. Member states allow either line-item rounding or invoice-total rounding, and the choice can change the final VAT by a cent or two per invoice. At scale that adds up. Companies operating across multiple EU countries usually pick one method and document it in their accounting policy memo. The Excel formula matters because the spreadsheet model has to match the policy.

Banking and financial services follow different rules again. ISO 20022 messaging standards specify rounding for currency conversions, often to four decimals for the exchange rate and two for the final amount. Reconciliation between counterparties depends on both sides using the same rounding logic โ€” if one bank uses ROUND and the other uses banker's rounding, settlement amounts diverge by fractions of a cent that, multiplied by millions of transactions, become real reconciliation breaks.

For most spreadsheet builders, the takeaway is simple. Pick the rounding function deliberately, document why, and round at the right boundary. Don't let display-format drift sneak into the value. Don't double-round. Test edge cases โ€” zero, negatives, large numbers โ€” before the model ships. The math is easy. The discipline is the hard part. Strong fundamentals here separate the analysts whose models tie out from the ones who spend Friday afternoon hunting penny variances.

Pick Your Rounding Function in 30 Seconds

Need standard rounding to N decimals? โ†’ ROUND(value, N)
Must never round down (shipping, billing units)? โ†’ ROUNDUP(value, N)
Must never round up (tax accruals, conservative)? โ†’ ROUNDDOWN(value, N)
Round to nearest 5, 25, or quarter hour? โ†’ MROUND(value, multiple)
Force up to next significance bracket? โ†’ CEILING.MATH(value, significance)
Force down to previous significance bracket? โ†’ FLOOR.MATH(value, significance)
Just drop the decimals, no rounding logic? โ†’ TRUNC(value)
Need mathematical floor (toward negative infinity)? โ†’ INT(value)
Round to even-only or odd-only integer? โ†’ EVEN(value) or ODD(value)
Need to compare floating-point sums? โ†’ wrap in ROUND(value, 10)

One Final Thought on Where Rounding Goes Wrong

Almost every rounding bug traces back to one of three causes. The first is treating cell format as if it changes value โ€” it doesn't, and SUM ignores formatting. The second is rounding too early in a calculation chain, which compounds error across downstream steps. The third is mismatched rounding logic across systems, where an Excel model rounds one way and the destination system (an ERP, a database, a regulatory filing) rounds another. None of these are math problems. They're discipline problems.

The practical fix is to follow three rules. Round at the display boundary, not before. Use the function whose behavior matches your business intent โ€” not the one that happens to be familiar. And document the rounding policy in a hidden tab or comment so the next person who opens the workbook (probably you, six months later) can see why ROUND was chosen over CEILING. That last one sounds like overkill until the day a colleague reopens the model, swaps the function "to be cleaner," and breaks six downstream reconciliations.

If you're new to spreadsheet work and want to drill these patterns into muscle memory, run a few example sheets โ€” round currency cells, round timestamps to 15 minutes, build a tax accrual that uses ROUNDDOWN, build a shipping table that uses CEILING. After ten or fifteen of these little exercises, the function names stop being abstract and the syntax becomes automatic. Going deeper into related areas like Excel functions and excel formulas rounds out the broader skillset, and a solid Excel cheat sheet keeps the most-used patterns one click away when memory fails.

Rounding looks trivial. It's not. The function-picking part takes one afternoon to learn. The discipline of using it correctly in production models takes longer. But once it's in your hands, every spreadsheet you touch gets more trustworthy โ€” and that, in the end, is the whole point of building models in the first place.

Test Your Excel Rounding Skills โ€” Free Practice

Excel Questions and Answers

How do I round a number in Excel?

Use the ROUND function: =ROUND(number, num_digits). For two-decimal currency, type =ROUND(A1, 2). For a whole integer, =ROUND(A1, 0). For the nearest hundred, =ROUND(A1, -2). The first argument is the value, the second tells Excel where the rounding cut-off sits โ€” positive for decimal places, zero for integers, negative for digits left of the decimal.

What's the difference between ROUND, ROUNDUP, and ROUNDDOWN?

ROUND uses standard arithmetic โ€” anything 0.5 or above rounds up, anything below rounds down. ROUNDUP always rounds away from zero, so 2.1 becomes 3 and -2.1 becomes -3. ROUNDDOWN always rounds toward zero, so 2.9 becomes 2 and -2.9 becomes -2. Pick by intent: ROUND for symmetric rounding, ROUNDUP when you can't underestimate, ROUNDDOWN when you can't overestimate.

How do I round to the nearest 5 or 10 in Excel?

Use MROUND. =MROUND(34, 5) returns 35. =MROUND(127, 10) returns 130. The function snaps your value to the nearest multiple of whatever you specify. The number and the multiple must share a sign โ€” both positive or both negative โ€” or MROUND returns a #NUM! error.

What does num_digits mean in the ROUND function?

num_digits is the second argument and it controls where the rounding cut-off sits relative to the decimal point. Positive numbers keep that many decimal places โ€” 2 means round to two decimals. Zero rounds to a whole integer. Negative numbers round to the left of the decimal โ€” -2 rounds to the nearest hundred, -3 to the nearest thousand.

Why does ROUND show 1.50 but SUM doesn't match?

Almost always a cell-formatting issue. Cell formatting changes the display but leaves the underlying value at full precision. If a cell shows 1.50 but actually holds 1.499, SUM uses 1.499 in the total. Fix by wrapping the source value in =ROUND(value, 2) so the stored value matches the display. Cell format is cosmetic; ROUND is surgical.

How do I round time to the nearest 15 minutes in Excel?

Use MROUND combined with TIME: =MROUND(A1, TIME(0,15,0)). Excel times are decimal fractions of a day, so 15 minutes is TIME(0,15,0) not literally 0.25. Using 0.25 by mistake rounds to quarter-days instead of quarter-hours, which produces some entertaining timesheet bugs.

What's the difference between INT and TRUNC in Excel?

INT rounds down toward negative infinity โ€” INT(-1.5) returns -2. TRUNC truncates toward zero โ€” TRUNC(-1.5) returns -1. For positive numbers both behave identically. The split matters with negatives, so check your data range before swapping one for the other in a formula.

Does Excel ROUND use banker's rounding?

No. Excel's ROUND function rounds half away from zero โ€” 2.5 becomes 3, -2.5 becomes -3. True banker's rounding (round-half-to-even) would send 2.5 to 2 and 3.5 to 4. R and Python's built-in round() default to banker's; Excel does not. If your team reconciles totals across these tools, document the rounding convention explicitly to avoid penny-level mismatches.
โ–ถ Start Quiz