Material Requirements Planning (MRP): How to Calculate What to Order, How Much, and When

The three inputs MRP requires

MRP is a dependent demand calculation. Independent demand (what end customers order) is handled by demand forecasting and the master production schedule. Dependent demand (what components are needed to build those end products) is calculated by MRP using three inputs.

The master production schedule (MPS) states what finished products will be produced, in what quantities, in which time periods. It is the demand signal that drives the entire MRP calculation. An MPS built from unrealistic demand forecasts or one that ignores capacity constraints will produce MRP output that cannot be executed.

The bill of materials (BOM) describes the structure of every product: which components go into each assembly, in what quantities, and at what level in the product hierarchy. A multi-level BOM traces from finished product through sub-assemblies down to purchased parts and raw materials. BOM accuracy is critical; a missing component or a wrong quantity factor means MRP will not plan replenishment for that item.

Inventory records contain the current on-hand quantity for every item, any scheduled receipts (open purchase orders or production orders already placed), and any quantities already allocated to existing orders. Inventory record accuracy is where most MRP implementations struggle: cycle counting (see the cycle counting post) is the primary tool for maintaining the record accuracy MRP requires.

The MRP calculation: netting, lot sizing, and lead time offsetting

The MRP calculation runs in three steps for each item at each level of the BOM.

Gross requirements are the total quantity of a component needed in each planning period, derived by multiplying the parent item’s planned production by the quantity per in the BOM.

Net requirements are gross requirements minus projected on-hand inventory and scheduled receipts. If projected on-hand exceeds gross requirements in a period, there is no net requirement; inventory is sufficient. If gross requirements exceed projected on-hand, the difference is a net requirement that must be covered by a planned order.

Planned orders are sized and timed to cover net requirements. Lot sizing rules determine order quantity: lot-for-lot ordering produces exactly what is needed each period (minimizing inventory but maximizing setup cost); fixed order quantity uses a predetermined batch size; period order quantity groups requirements across multiple periods. After sizing, the planned order release date is calculated by offsetting backward from the required date by the item’s planned lead time.

BOM levels and low-level coding

In a multi-level product structure, the same component may appear at multiple levels of multiple products. Low-level coding assigns each item the lowest level at which it appears anywhere in any BOM. MRP processes level by level from the top of the BOM down, and low-level coding ensures a component is not planned until all higher-level demand for it has been calculated. Without low-level coding, MRP would generate incorrect net requirements for shared components.

Time fencing and the planning horizon

The MRP planning horizon must extend at least as far as the longest cumulative lead time in the product structure: the sum of lead times from raw material procurement through every level of the BOM to the finished product. If the planning horizon is shorter than the cumulative lead time, MRP cannot generate orders early enough to support the master schedule.

Time fences divide the planning horizon into zones. Inside the demand time fence (closest in), the MPS is frozen: changes require management authorization because production is already committed or in process. Between the demand time fence and the planning time fence, changes are allowed but must go through a formal change process. Beyond the planning time fence, the MPS is a forecast-driven plan open to automatic adjustment.

MRP nervousness and how to manage it

MRP nervousness describes the tendency of MRP systems to reschedule orders in response to small changes in demand or supply. A one-unit change in the MPS can cascade through multiple BOM levels, generating rescheduling messages for dozens of orders. Nervousness is not a defect in the logic; it is the correct mathematical response to changes. The problem is that not every rescheduling message represents a real execution problem worth acting on.

Managing nervousness requires: frozen time fences that protect near-term production from constant replanning; safety stock at selected levels to absorb minor demand variation without triggering replanning; and planner judgment to filter rescheduling messages and act only on those that represent genuine execution risks.

MRP II: closing the loop to capacity

Classic MRP is a material planning system; it calculates what to order but does not check whether the production orders it generates are actually feasible given available capacity. Manufacturing resource planning (MRP II) extended MRP by closing this loop: MRP II feeds planned orders into capacity requirements planning (CRP), which checks workcenter loads, and then feeds the results back to either confirm the MPS is feasible or flag capacity violations that require MPS adjustment. This feedback loop between material planning and capacity planning is what distinguishes MRP II from the original MRP logic.

MRP versus reorder point systems

Reorder point (ROP) systems replenish when on-hand inventory falls to a predetermined level, based on historical average demand and lead time. MRP calculates replenishment from forward-looking demand (the MPS) rather than backward-looking consumption. For independent demand items with stable, continuous consumption, ROP systems are simpler and often adequate. For dependent demand items in manufacturing, where demand is lumpy and driven by scheduled production, MRP produces more accurate replenishment timing and typically lower average inventory than ROP systems.

Common mistakes

• Running MRP with inaccurate inventory records. Inventory record accuracy below 95 percent produces MRP output that planners cannot trust. Cycle counting is the prerequisite for effective MRP, not an optional best practice.
• Using BOMs that do not reflect current product structure. Engineering changes not reflected in the BOM cause MRP to plan for the wrong components. BOM maintenance must be treated as a real-time operational process, not a periodic cleanup task.
• Setting lead times that do not match reality. MRP uses planned lead times to offset order releases. If planned lead times are optimistic, orders are released too late. If they are padded conservatively, inventory inflates. Lead times in MRP should reflect actual performance, reviewed periodically.
• Treating every MRP rescheduling message as an action item. MRP generates rescheduling messages for any timing mismatch. Acting on all of them creates churn without improving service. Planners need clear criteria for which messages represent real execution risks versus noise.

Frequently asked questions

What is material requirements planning (MRP)?

Material requirements planning (MRP) is a time-phased calculation that determines what materials and components are needed, in what quantities, and when they must be available to support a production schedule. It takes the master production schedule as the demand signal, explodes it through the bill of materials to calculate gross requirements for every component at every level, nets out on-hand inventory and scheduled receipts to find what actually needs to be ordered, applies lot sizing rules to determine order quantities, and offsets by lead time to set order release dates. The output is a set of planned orders for purchased materials and manufactured components.

What are the three main inputs to MRP?

The three inputs are: (1) the master production schedule (MPS), which states what finished products will be produced and when; (2) the bill of materials (BOM), which describes the structure of each product and the quantity of each component required; and (3) inventory records, which contain current on-hand quantities, scheduled receipts from open orders, and allocated quantities. The quality of MRP output depends directly on the accuracy of these three inputs. An inaccurate inventory record causes MRP to calculate wrong net requirements; an incorrect BOM causes it to plan for the wrong components; an unrealistic MPS causes it to generate orders that cannot be executed.

What is the difference between gross requirements and net requirements in MRP?

Gross requirements are the total quantity of a component needed in a time period, calculated by multiplying the parent item’s planned production quantity by the component’s quantity-per in the BOM. Net requirements are what actually needs to be ordered: gross requirements minus projected on-hand inventory and any scheduled receipts already inbound. If projected on-hand covers the gross requirement, there is no net requirement and no order is needed. If gross requirements exceed projected on-hand, the difference is a net requirement that triggers a planned order. The netting step is what allows MRP to avoid ordering material that is already available.

What is MRP nervousness and how do you manage it?

MRP nervousness is the tendency of MRP systems to generate frequent rescheduling recommendations in response to small changes in demand or supply. Because MRP recalculates the entire plan each time it runs, even minor changes in the MPS can cascade through the BOM and produce rescheduling messages for many orders. Managing nervousness involves three tools: (1) frozen time fences that prevent MPS changes within the execution horizon without management authorization; (2) safety stock at selected levels to absorb minor variation without triggering replanning; and (3) planner judgment to filter action messages and act only on those representing genuine service risks. Nervousness itself is not a flaw; the flaw is acting on every message without filtering.

What is the difference between MRP and MRP II?

MRP (material requirements planning) calculates what materials and components are needed and when, based on the master production schedule and BOM. It is a material planning system only; it does not check whether the production orders it generates are feasible given available capacity. MRP II (manufacturing resource planning) closes this loop by feeding planned production orders into capacity requirements planning (CRP), which checks workcenter loads against available capacity. When CRP identifies a capacity violation, the MPS can be adjusted, and MRP recalculates. MRP II also typically integrates financial planning, so production plans tie directly to costs and budgets. Enterprise resource planning (ERP) systems evolved from MRP II by extending the same closed-loop logic across the full business.

When should you use MRP versus a reorder point system?

Use MRP when demand is dependent (driven by a production schedule) and lumpy: for manufactured components and purchased materials whose consumption is tied to discrete production runs rather than continuous end-customer demand. MRP calculates replenishment from forward-looking demand signals and handles variable demand patterns better than reorder point systems. Use reorder point (ROP) systems for independent demand items with relatively stable, continuous consumption, where the simplicity of a min-max or two-bin system outweighs the overhead of MRP. Many operations use both: MRP for the primary production BOM, and ROP or min-max for MRO supplies and low-value consumables that are not worth including in the MRP calculation.