
What is the master production schedule (MPS)
The master production schedule (MPS) states how much of each finished product will be built in each period of the planning horizon, usually weeks. It turns the aggregate sales and operations plan into quantities and dates, and feeds MRP, which calculates components and materials.
In practice the MPS answers one question: given my forecast, booked orders, inventory and capacity, what will I build and when? It is the central link of production planning and control.
Visual map of the master production schedule
The map shows where the MPS sits between S&OP and MRP and the period table of the worked example below.

Where the MPS fits
- S&OP (aggregate plan): sets volumes per product family, in units or hours, months ahead.
- MPS: breaks the family into finished products by week, respecting inventory and capacity.
- MRP: explodes the MPS through the bill of materials and creates purchase and production orders.
- Execution: sequencing and shop floor control, as in production sequencing.
Inputs are the demand forecast, firm orders, beginning inventory, lot size and safety stock.
Worked example: 8 weeks
Data: beginning inventory of 50 units, fixed lot of 100, zero safety stock. In each week, demand is the larger of forecast and firm orders. Projected on-hand (POH) is:
POH(t) = POH(tβ1) + MPS(t) β demand(t); schedule a lot when POH(tβ1) β demand(t) < 0
Week 1: 50 β 70 = β20, negative, so a lot of 100 is scheduled and POH becomes 50 + 100 β 70 = 80. Week 2: 80 β 60 = 20, no lot. Week 3: 20 β 70 = β50, a lot of 100 is scheduled and POH is 50. The full result:
| Item | Wk 1 | Wk 2 | Wk 3 | Wk 4 | Wk 5 | Wk 6 | Wk 7 | Wk 8 |
|---|---|---|---|---|---|---|---|---|
| Forecast | 60 | 60 | 70 | 70 | 80 | 80 | 90 | 90 |
| Firm orders | 70 | 55 | 40 | 25 | 10 | 0 | 0 | 0 |
| Demand used | 70 | 60 | 70 | 70 | 80 | 80 | 90 | 90 |
| MPS (lot 100) | 100 | β | 100 | 100 | β | 100 | 100 | 100 |
| Projected on-hand | 80 | 20 | 50 | 80 | 0 | 20 | 30 | 40 |
| ATP | 25 | β | 60 | 65 | β | 100 | 100 | 100 |
Check: 50 beginning + 600 produced β 610 demand = 40, equal to week 8 inventory. Six lots of 100 cover the eight weeks, averaging 75 per week against an average demand of 76.25.
ATP: how much can I promise a customer
ATP (Available to Promise) is the part of the MPS not yet committed to orders, up to the next lot. In the first week it includes beginning inventory.
ATP(week 1) = beginning inventory + MPS β orders until next MPS; other weeks: MPS β orders until next MPS
Week 1: 50 + 100 β (70 + 55) = 25. Week 3: 100 β 40 = 60. Week 4: 100 β (25 + 10) = 65. Weeks 6, 7 and 8 have no orders and keep 100 each. If sales wants to promise 80 units for week 3, the ATP of 60 shows the lot must be moved earlier or another one created.
Capacity check (RCCP)
The MPS is only valid if it fits capacity. The rough-cut check uses the bottleneck resource. Assume 0.2 h per unit at the bottleneck and 25 h available per week:
| Item | Wk 1 | Wk 2 | Wk 3 | Wk 4 | Wk 5 | Wk 6 | Wk 7 | Wk 8 |
|---|---|---|---|---|---|---|---|---|
| MPS | 100 | β | 100 | 100 | β | 100 | 100 | 100 |
| Bottleneck hours | 20 | 0 | 20 | 20 | 0 | 20 | 20 | 20 |
| Load | 80% | 0% | 80% | 80% | 0% | 80% | 80% | 80% |
Each lot of 100 takes 100 Γ 0.2 = 20 h, or 20 Γ· 25 = 80% of capacity. Average load is (600 Γ 0.2) Γ· (8 Γ 25) = 120 Γ· 200 = 60%. If a lot exceeded 25 h, you would need to split it, build part earlier or adjust capacity. See how to calculate production capacity and constraints in the theory of constraints.
Horizon, time fences and leveling
- Horizon: should cover at least the product's cumulative lead time so MRP sees the longest purchases.
- Time fences: in the first weeks the MPS is frozen or changes only with approval; further out, the planner adjusts more freely.
- Forecast vs orders: near the date, firm orders rule; far out, the forecast rules. The example uses the larger of the two for simplicity.
- Leveling: very uneven lots overload the plant. Heijunka smooths mix and volume, and safety stock protects against variation.
For inventory sizing, see inventory management.
Common mistakes
- Ignoring capacity and producing an MPS the shop floor cannot meet.
- Changing the MPS every week with no time fence, creating MRP nervousness.
- Planning at component level instead of finished product or planning item.
- Wrong inventory base: system balance differing from the physical count.
- Promising orders without checking ATP, overcommitting full lots.
- Unreviewed forecast with no error tracking.
Frequently asked questions
What is the MPS in manufacturing?
The master production schedule: quantities of finished products to build in each period of the horizon.
What is the difference between MPS and MRP?
The MPS sets what and how much finished product to build; MRP calculates the components and materials needed to meet it.
How do I calculate projected on-hand in the MPS?
POH(t) = POH(tβ1) + MPS(t) β demand(t). When the result would be negative, schedule a lot.
What is ATP?
Available to Promise: the MPS minus the orders already committed until the next lot.
What are time fences?
Time windows in which the MPS does not change, or changes only with approval, to keep production and purchasing stable.
How long should the MPS horizon be?
At least the product's cumulative lead time. In many plants, 8 to 12 weeks.
Sources
- VOLLMANN, T. E. et al. Manufacturing Planning and Control for Supply Chain Management. McGraw-Hill.
- ASCM (APICS). Dictionary.
- ORLICKY, J. Material Requirements Planning. McGraw-Hill, 1975.
- STEVENSON, W. J. Operations Management. McGraw-Hill.
Want a quick reference guide for the factory floor?
Download the free Pocket Guide e-book with the Lean tools summarized on one page each.
Download the e-book

