The weighted average cost method pools the costs of similar inventory and assigns one average cost per unit. Divide total cost of goods available for sale by total units available, then multiply that average by units sold and units remaining. In a perpetual system, recalculate the moving average after each purchase.
Weighted average cost per unit = total cost of goods available for sale ÷ total units available for sale
—
—
COGS = weighted average cost per unit × units sold
—
—
Ending inventory = weighted average cost per unit × units remaining
—
—
Beginning inventory contributes both units and dollars. Do not average its per-unit cost with purchase prices without weighting by quantity. A layer of 200 units must affect the average more than a layer of 20 units.
Net purchase cost includes the costs that the applicable accounting policy treats as bringing inventory to its present location and condition. Use the full cost pool supplied by the problem.
Units available must use the same scope as total cost. Dividing the cost of three layers by the units in only two layers produces an average that cannot reconcile.
Units sold and remaining split the pool after the average is calculated. Both receive the same periodic average unit cost.
A retailer has 450 identical desk lamps available in July: 120 units at $18.40, 180 at $19.75, and 150 at $21.10. Total cost is $8,928. The retailer sells 320 lamps, leaving 130.
The weighted average is $19.84 per lamp because $8,928 divided by 450 is exactly $19.84. Apply that rate to both sides of the split. COGS is $6,348.80 and ending inventory is $2,579.20.
| Step | Calculation | Result |
|---|---|---|
| First cost layer | 120 × $18.40 | $2,208.00 |
| Second cost layer | 180 × $19.75 | $3,555.00 |
| Third cost layer | 150 × $21.10 | $3,165.00 |
| Weighted average unit cost | $8,928 ÷ 450 | $19.84 |
| Cost of goods sold | 320 × $19.84 | $6,348.80 |
| Ending inventory | 130 × $19.84 | $2,579.20 |
| Check | $6,348.80 + $2,579.20 | $8,928.00 |
If the 320 lamps sold for $31 each, revenue is $9,920 and gross profit is $3,571.20. Weighted average affects the cost assignment, not the selling price or quantity sold.
The periodic method computes one average after combining beginning inventory and all purchases for the period. It applies that one rate to every unit sold and remaining.
A perpetual system uses a moving average. After a purchase, combine the cost and quantity already on hand with the new purchase, then calculate a new per-unit cost. A sale uses the latest average established before that sale. A sale reduces units and dollars at the current average but does not itself create a new average.
For example, begin with 120 lamps costing $2,208, or $18.40 each. Buy 180 lamps for $3,555. The new average is $19.21 because $5,763 divided by 300 units is $19.21. If 100 lamps then sell, COGS is $1,921 and 200 units costing $3,842 remain. A later purchase of 150 units for $3,165 creates a new average of $20.02: $7,007 divided by 350. No rounding adjustment is needed for that second average.
Because periodic weighted average waits until period-end while moving average responds to purchase timing, the two methods can assign different costs when purchases and sales alternate. Always identify the inventory system before calculating.
Do not round the average too early unless the problem instructs you to. Keep extra decimal places in the unit cost, calculate COGS and ending inventory, then round the reported dollar amounts. If both amounts are independently rounded, a small difference can remain. Assign any stated rounding adjustment consistently so COGS plus ending inventory equals total cost available.
The examples avoid that issue because both $19.84 and $20.02 are exact. Other data sets can produce repeating decimals, so retain the unrounded quotient until the reported amounts are calculated.
Weighted average is useful when units are ordinarily interchangeable and tracking separate historical layers adds little information. IAS 2 permits weighted average for such inventory, as does U.S. GAAP. The method smooths differences among purchase prices, so its COGS and ending inventory often fall between FIFO and LIFO when prices change steadily.
It is not the same as specific identification. A dealer may trace a particular vehicle's exact cost, while a fuel distributor may pool indistinguishable units. The nature of the inventory and the accounting framework determine which formula is appropriate.
Taking a simple average of prices. The average of $18.40, $19.75, and $21.10 is not the inventory average unless each layer has the same quantity. Multiply each price by its units first.
Using units sold as the denominator. The rate uses all units available for sale, not only the units sold.
Mixing periodic and moving average. Periodic uses one end-of-period rate. Moving average recalculates after purchases.
Recalculating after a sale. A sale uses the current moving-average rate. It changes the size of the pool, not the per-unit average.
Ignoring the dollar check. COGS plus ending inventory must equal total cost available, subject only to an explained rounding adjustment.
A perpetual worksheet needs five running figures after each transaction: units on hand, inventory dollars on hand, average unit cost, units sold, and COGS. A purchase changes units and dollars, so divide the new dollar balance by the new quantity. A sale uses that current unit cost to reduce both balances.
Suppose 200 lamps costing $3,842 remain after a sale, so their average is $19.21. The later purchase of 150 lamps for $3,165 raises the pool to 350 units costing $7,007. The exact new rate is $7,007 divided by 350, or $20.02 when displayed to cents. If 80 lamps sell next, calculate COGS using the unrounded rate stored by the system, then leave 270 units at the same unrounded average. The sale does not blend in a new cost.
This running ledger also catches negative inventory errors. A system cannot assign a sensible moving average if it records a sale before the related stock is on hand. In an exercise, negative units usually signal that transactions were processed out of order. In practice, timing and data errors should be corrected rather than hidden by a later purchase.
For statement analysis, weighted average dampens the effect of a single unusually high or low purchase compared with a pure layer method. That smoothing does not make the method more accurate for every inventory. It is appropriate when pooling interchangeable units represents the cost flow selected under the entity's accounting policy.
Weight every unit cost by its quantity: total pool cost divided by total pool units. Use one rate for periodic weighted average, but recalculate after each purchase under perpetual moving average.
Divide total cost of goods available for sale by total units available for sale. Multiply the resulting unit cost by units sold for COGS and by units remaining for ending inventory.
In inventory accounting, the terms commonly describe the same pooling approach. The average must be weighted by quantities, not found by taking a simple mean of purchase prices.
Moving average is the perpetual version of average costing. The business recalculates unit cost after each purchase and uses the latest average for the next sale.
Not always. When purchases and sales alternate, periodic weighted average uses all period costs together, while moving average uses only costs available before each sale.
Yes. IAS 2 permits the weighted average cost formula for ordinarily interchangeable inventory.
Carry extra decimal places in the average until the final amounts. Then confirm COGS plus ending inventory equals total cost available and handle any small rounding difference consistently.
By the FinanceBrain Team · Last verified July 12, 2026 · How we produce and verify articles