Dividing by a fraction answers the question "how many groups of this fraction fit into that amount?" 3 / (1/4) = 12, because 12 quarter-pieces fit into 3 wholes. The algorithm -- "invert and multiply" (multiply by the reciprocal) -- is efficient but needs conceptual grounding. Dividing a/b by c/d is equivalent to multiplying a/b by d/c because division asks "how many of c/d fit in a/b?" and flipping the divisor converts the question into a multiplication. At fifth grade, students focus on dividing a whole number by a unit fraction and a unit fraction by a whole number, with extension to general fraction division.
Start with whole numbers divided by unit fractions using visual models: "How many 1/4-cup servings are in 3 cups?" Draw 3 rectangles, partition each into 4 parts, count 12. Then do unit fractions divided by whole numbers: "If 1/2 a pizza is shared among 3 people, each gets 1/6." Build toward the invert-and-multiply rule, always grounding it in the question "how many groups?"
Division always asks the same question: how many groups of this size fit into that total? 12 ÷ 3 asks "how many groups of 3 fit into 12?" — the answer is 4. When you divide by a fraction, the question is identical, just with a fractional group size. 3 ÷ (1/4) asks "how many quarter-sized groups fit into 3 wholes?" Picture three rectangles, each cut into four equal pieces — you have 12 quarter-pieces in total. So 3 ÷ (1/4) = 12. Notice that dividing by a number *less than 1* gave an answer *larger* than what you started with. This surprises many students who assume division always makes things smaller — but that assumption only holds when you divide by numbers greater than 1.
The invert-and-multiply rule — "flip the divisor and multiply" — is a shortcut for this counting process. To divide a/b by c/d, you compute (a/b) × (d/c). The reason it works is this: multiplying by the reciprocal d/c is precisely equivalent to asking "how many times does c/d fit?" You already know how to multiply fractions from your prerequisite, so once you accept the rule you can execute it reliably. The critical step is to flip the *divisor* (the number you are dividing *by*), not the dividend (the number being divided). A common error is flipping the wrong one. Keep track by labeling: "What am I dividing by?" — that is the fraction to flip.
For unit fractions divided by whole numbers — the other main case at this level — the picture is different. (1/2) ÷ 3 asks "if 1/2 of a pizza is shared equally among 3 people, how much does each person get?" Draw half a pizza and cut that half into 3 equal slices. Each slice is 1/6 of the whole pizza. So (1/2) ÷ 3 = 1/6. Using the rule: flip 3 to get 1/3, then (1/2) × (1/3) = 1/6. The result is smaller than the original fraction, which makes sense — you are splitting a part into even smaller parts. Visualizing both cases — a whole divided into fraction-sized pieces (result grows) and a fraction divided into whole-number groups (result shrinks) — builds the conceptual intuition that no amount of practice with the algorithm alone can provide.