The Complete Guide to Multi-Digit Multiplication

The homework says 36 × 94, and your child has drawn a box with four smaller boxes inside it. You learned to stack and carry. Both ways reach the same answer, and each has its place.
This guide takes one problem through the four methods you are likely to see: partial products, the area model, the traditional algorithm, and the lattice. Every step is worked out, so you can follow along with a pencil and help your 4th or 5th grader tonight.
What Kids Are Expected to Do, and When
In the Common Core, multi-digit multiplication starts in 4th grade. Under standard 4.NBT.B.5, kids work up to four digits times one digit and two digits times two digits, leaning on place value and the properties of operations as they go.
They also have to illustrate and explain each calculation, with equations, arrays, or area models. That is why the boxes show up on homework. Teachers are asking students to show why the answer is right, not only to produce it.
In 5th grade, standard 5.NBT.B.5 takes one line: kids should be fluent with the standard algorithm. So the method you probably learned is not gone. It comes after the others, and in states with their own standards, the timing may differ.
The authors of the standards explain the order in their Progressions document. Kids are meant to use methods they can make sense of and explain, and over time the algorithms become abbreviations or summaries of their reasoning. The boxes come first so that the stacked version makes sense later.
What Every Method Rests On
All four methods depend on the same two skills, so it is worth checking them before blaming the method.
The first is the basic facts. Inside 36 × 94 are four of them: 3 × 9, 3 × 4, 6 × 9, and 6 × 4. If a child is still counting up to find 6 × 9, the big problem becomes a long and error-prone climb. The 9 times table page and its neighbors are the place to shore up any fact that slows things down.
The second is multiplying by tens and hundreds. The standards guide gives this example: 6 groups of 7 hundreds is 6 × 7 hundreds, which is 42 hundreds, or 4,200. Likewise, 30 × 90 is 3 tens times 9 tens, or 27 hundreds, which is 2,700. If that thinking feels shaky, go back to it before going further.
In American notation, 36 × 94 means 36 groups of 94. Since the order does not change the answer, you can also picture it as 94 groups of 36, and the methods below work either way.
Partial Products
Partial products means breaking both numbers into place value parts, multiplying every part by every part, and adding the results. Nothing is hidden, and each line shows what it stands for.
One-Digit Example: 8 × 549
Split 549 into 500, 40, and 9. Then multiply each part by 8:
- 8 × 500 = 4,000
- 8 × 40 = 320
- 8 × 9 = 72
Add them: 4,000 + 320 + 72 = 4,392. You can work from left to right, as here, or from right to left. Either way, the answer is the same, and starting with the largest part gives a good early sense of the size of the answer.
Two-Digit Example: 36 × 94
Split 36 into 30 and 6, and 94 into 90 and 4. Now there are four multiplications:
- 30 × 90 = 2,700
- 30 × 4 = 120
- 6 × 90 = 540
- 6 × 4 = 24
Add them: 2,700 + 120 is 2,820, plus 540 is 3,360, plus 24 is 3,384. A quick estimate confirms it: 40 × 90 is 3,600, and 36 is a bit less than 40, so an answer a little below 3,600 makes sense.
The Area Model
The area model is partial products with a picture. Draw a rectangle and imagine its sides are 36 and 94. Its area is the product. Cut the top edge into 90 and 4, and the side edge into 30 and 6. That cuts the rectangle into four smaller boxes.
Fill in each box with the area of that piece. The top row holds 30 × 90 = 2,700 and 30 × 4 = 120. The bottom row holds 6 × 90 = 540 and 6 × 4 = 24. Add all four boxes, and you get 3,384, the same four numbers as before.
The picture makes one thing very clear: every part of one number gets multiplied by every part of the other. Write only 30 × 90 and 6 × 4, and two empty boxes stare back at you. The drawing does not need to be to scale, so a quick sketch is fine.
This model also grows. For 243 × 57, the rectangle has three columns (200, 40, and 3) and two rows (50 and 7), which makes six boxes: 10,000, 2,000, 150, 1,400, 280, and 21. They add up to 13,851.
The Traditional Algorithm
Many parents remember this stacked method. It records the same multiplications in a compact form, and that compact form is both its strength and its danger. Here are the steps for 36 × 94, with 94 written on top and 36 under it, lined up on the right.
- Multiply the top number by the 6. First 6 × 4 = 24: write the 4 in the ones place and carry the 2. Next 6 × 9 = 54, plus the carried 2 is 56. The first line reads 564.
- Start a second line for the 3 in 36, which really means 30. Write a 0 in the ones place.
- Multiply by the 3. First 3 × 4 = 12: write the 2 in the tens place and carry the 1. Next 3 × 9 = 27, plus the carried 1 is 28. The second line reads 2,820.
- Add the two lines: 564 + 2,820 = 3,384.
That zero in step 2 is the part kids most need explained. The Progressions document says it is there because the whole line is produced by multiplying by 30, not 3. Compare the lines with the area model: 564 is 540 + 24, and 2,820 is 2,700 + 120. The boxes are still there, just folded together.
Carried digits need care, too. The same document warns that placing those digits above the 94 can put them in incorrect places. Some teachers have students write them small, and the document says that minor variations in how an algorithm is recorded are acceptable. When your child's teacher writes it differently from the way you do, the teacher's way is fine.
The steps stay the same for a three-digit number. For 243 × 57, 7 × 243 gives 1,701, and 50 × 243 gives 12,150. The sum is 13,851, matching the six boxes above.
The Lattice Method
Lattice multiplication uses a grid with a diagonal line through every square. It looks modern, but it is old. Historical accounts trace it back to at least the late 1200s and early 1300s, and it was printed in Italy in the 1400s. Its Italian name, gelosia, refers to a window grating, and that is what the grid resembles.
The Common Core does not mention it, so some classes will use it and others never will. Here it is for 36 × 94. Write 9 and 4 across the top of a two-by-two grid, and 3 and 6 down the right side. Draw a diagonal from the upper right to the lower left of each square.
Fill each square with one fact, the tens digit above its diagonal and the ones digit below it. The top row holds 27 and 12, for 3 × 9 and 3 × 4. The bottom row holds 54 and 24, for 6 × 9 and 6 × 4. All the multiplying is now done, and nothing has been carried yet.
Next, add along the diagonal lanes, starting at the lower right. The first lane holds only the 4. The next holds 2, 4, and 2, which make 8. The next holds 5, 7, and 1, which make 13, so write 3 and carry 1 into the last lane. That lane holds 2 plus the carried 1, which makes 3. Read the digits from the left: 3,384.
A 1973 study of 264 4th graders, later published in a journal, found that those taught the lattice were more accurate and faster than those taught a distributive method, with no difference in understanding or attitude. It is one dated study, so treat it as a sign that the lattice can work for some children and not as a reason to switch.
Which Method Should Your Child Use?
Use the one the class is using right now. Mixing methods at home during a new unit can confuse a child, and the teacher may be asking for a specific one on tests. Once it is solid, a second method becomes a handy way to check answers.
Partial products and the area model suit a child who is learning why multiplication works, or one who loses track when numbers get folded together. They take more writing, and more room on the page. The traditional algorithm is quicker once it is understood, and it is the one named in the 5th grade standard.
The lattice suits a child who knows the facts but gets tangled in carrying, since the carrying waits until the end. Its weakness is that the grid hides place value, so ask your child what the 27 in the top left square really stands for. (It is 30 × 90, or 2,700.)
Fixing the Usual Errors
When an answer is wrong, look for a pattern before you reteach anything. A report from 1974 on 744 children in grades 2 to 6 found that 5 to 6 percent made systematic errors in the addition, multiplication, and division algorithms, meaning they made the same wrong step again and again. A year later, 23 percent of the children were making the same error or another systematic one.
That report is old, but its message still helps: a repeated mistake is often a misunderstanding, and more worksheets alone may not clear it up. Have your child talk through one problem out loud while you listen, and you can usually hear where things go off the track.
The errors below come up in the steps above, and each one has a simple check:
- Forgetting the zero on the second line. Ask what the 3 in 36 is worth, then have your child compare the line with the area model. A much-too-small answer, such as 846 for 36 × 94, is the telltale sign.
- Adding the carried digit before multiplying. In 6 × 94, the 2 is added after 6 × 9, not before. Say each step aloud: multiply, then add what you carried.
- Using the old carry on the new line. Cross off carried digits when a line is done, so a leftover 2 does not sneak into the next one.
- Misaligned columns. Turn lined paper sideways so the lines become columns, or use graph paper with one digit per square.
- A wrong basic fact. If the steps are right and the answer is off, check each fact. A single 6 × 9 = 56 throws off the entire problem.
An estimate catches many of these errors. Before starting, round the numbers and multiply. If 40 × 90 says about 3,600 and the worked answer is 846 or 33,840, your child knows at once that something went wrong and can look for it.
A few problems done carefully, with one explained out loud, will do more than a full page rushed. When the steps are clear but the facts are slow, a short daily round of fact practice is a good place to start. See what fact fluency really means for how to judge that.
