The predetermined overhead rate (POHR) is estimated total manufacturing overhead for the year divided by the estimated total amount of an activity driver, such as direct labor hours or machine hours. It is set before the year starts, then used all year to apply overhead to jobs as the work happens. Because both inputs are estimates, applied overhead rarely equals actual overhead, and the difference is closed out at year-end as over- or underapplied overhead.
POHR = Estimated total manufacturing overhead ÷ Estimated total activity driver
POHR — Predetermined overhead rate, in dollars per unit of the driver (e.g. $ per direct labor hour)
Estimated total manufacturing overhead — Budgeted overhead cost for the coming year — indirect materials, indirect labor, rent, utilities, depreciation
Estimated total activity driver — Budgeted amount of the allocation base for the same year — direct labor hours, machine hours, or direct labor cost
Applied overhead = POHR × Actual driver activity
Applied overhead — The overhead cost assigned to jobs during the year
Actual driver activity — The driver amount actually used — actual hours worked, not the estimate
Both numbers in the rate are estimates. That is the whole point of the word predetermined: the rate exists before a single actual overhead bill arrives.
Overhead costs arrive late and lumpy. The December utility bill shows up in January. The factory insurance premium hits once, in March. A machine breakdown produces a repair invoice in whatever month it happens to fail. If you waited for actual overhead, you could not cost a job until well after year-end.
Jobs cannot wait that long. A cabinet shop that finishes a custom order on February 9 needs the full cost of that job now — to bill the customer, to quote similar work, and to carry the job in inventory at a defensible number. In job-order costing, overhead is the one product cost you cannot trace directly to a job, so the POHR stands in: it converts a year of messy, delayed overhead bills into a steady per-hour charge that can go on a job cost sheet the day the job finishes.
An annual rate also smooths out seasonality. If overhead were assigned month by month at actual cost, two identical jobs would carry different amounts of heating cost just because one ran in January and one ran in July. One rate for the whole year keeps identical jobs at identical costs.
The denominator is not a free choice. The driver should be the activity that causes overhead to be incurred — a cause-and-effect link, not just something easy to count.
In a labor-intensive shop, most overhead (supervision, payroll support, workspace costs) rises and falls with people-hours, so direct labor hours or direct labor cost is the natural base. In a highly automated plant, overhead is dominated by machine depreciation, power, and maintenance, so machine hours track it far better. Some single-product or simple operations use units produced.
The choice matters because a mismatched driver misprices jobs. If a plant runs on machines but applies overhead by labor hours, a hand-finished job that barely touches the machines absorbs a large share of machine-driven overhead it never caused, while machine-heavy jobs look artificially cheap. Exam questions test this idea by asking which base is most appropriate for a described operation; the answer is always the one with the strongest causal link to the overhead pool.
| Step | Computation | Result |
|---|---|---|
| Estimated manufacturing overhead (set in December, before the year) | From the annual budget | $391,500 |
| Estimated direct labor hours | From the annual budget | 27,000 DLH |
| Predetermined overhead rate | $391,500 ÷ 27,000 DLH | $14.50 per DLH |
| Actual direct labor hours worked during the year | From payroll records | 26,300 DLH |
| Overhead applied to jobs | 26,300 DLH × $14.50 | $381,350 |
| Actual overhead incurred | From the year's actual bills | $389,740 |
| Difference | $389,740 − $381,350 | $8,390 underapplied |
Read the last three rows carefully. The company charged jobs $381,350 of overhead through the rate, but the real bills totaled $389,740. Jobs absorbed $8,390 too little cost, so overhead is underapplied by $8,390. If applied overhead had come out above $389,740 instead, it would be overapplied.
Because the POHR is built from estimates, a year-end gap between applied and actual overhead is normal, not a mistake.
Underapplied means applied overhead is less than actual overhead. Jobs were charged too little; product costs on the books are understated. Overapplied means applied overhead is more than actual overhead. Jobs were charged too much; product costs are overstated.
The simplest treatment — and the default in most intro courses — closes the entire difference to Cost of Goods Sold:
The logic is direction-checking, not memorization. Underapplied means the expense charged through jobs was too small, so the close-out adds the missing expense back; more expense means less income. Overapplied is the mirror image. (A more precise method prorates the difference across Work in Process, Finished Goods, and COGS, but the close-to-COGS shortcut is what most homework asks for.)
Mixing estimated and actual in the rate. The POHR is estimated overhead over estimated driver — both numbers come from the budget, before the year starts. If a problem hands you actual overhead of $389,740 and estimated hours of 27,000, dividing them produces a number that is neither the POHR nor an actual rate. Match estimate with estimate.
Applying overhead with the estimated driver. The denominator of the rate is estimated hours, but application uses actual hours: $14.50 × 26,300, not $14.50 × 27,000. Multiplying the rate by estimated hours just gives back the overhead budget and ignores what actually happened on the shop floor.
Flipping the over/under direction. Students see "actual exceeded applied" and guess overapplied because actual is the bigger number. The label describes applied overhead: applied too little → underapplied; applied too much → overapplied. Anchor on applied, then compare to actual, and the income effect follows — underapplied raises COGS and lowers income, overapplied does the opposite.
The predetermined overhead rate is estimated overhead divided by an estimated driver, fixed before the year begins; overhead is applied at that rate times actual driver activity, and the year-end gap between applied and actual overhead closes to COGS — underapplied raises COGS and lowers income, overapplied does the reverse.
Divide estimated total manufacturing overhead for the year by the estimated total amount of the activity driver, such as direct labor hours or machine hours. Both numbers come from the budget set before the year starts. For example, $391,500 of estimated overhead over 27,000 estimated direct labor hours gives a rate of $14.50 per direct labor hour.
Actual overhead arrives late and unevenly — insurance premiums hit once a year, repair bills come whenever machines fail, and utility costs swing with the seasons. Jobs need costs as soon as they finish, for billing, quoting, and inventory. A predetermined rate lets a company assign overhead the moment work happens instead of waiting for the actual bills, and it keeps identical jobs at identical costs year-round.
Actual hours. Only the rate itself is built from estimates. During the year, applied overhead equals the predetermined rate times the driver activity that actually occurred — for example, $14.50 per hour times 26,300 actual direct labor hours, not the 27,000 hours that were budgeted.
Applied overhead exceeded actual overhead, so jobs were charged more overhead cost than the company really incurred. Under the common close-to-COGS treatment, the overapplied amount is subtracted from Cost of Goods Sold, which lowers COGS and raises net operating income for the period.
The one with the strongest cause-and-effect link to the overhead pool. Labor-intensive operations usually use direct labor hours or direct labor cost, because overhead there rises with people-hours. Automated plants usually use machine hours, because depreciation, power, and maintenance track machine time. A poorly matched driver systematically misprices jobs.
A plantwide rate is one kind of predetermined rate — a single rate covering the whole factory. Companies whose departments differ a lot (one automated, one manual) often compute a separate predetermined rate per department, each with its own overhead estimate and its own driver, to keep the cause-and-effect link tight.
By the FinanceBrain Team · Last verified July 11, 2026 · How we produce and verify articles