AACE Estimate Classes
1 Through 5 Explained

Why Estimates Are Classified

AACE International published recommended practice 18R-97 to give the industry a shared vocabulary for estimate reliability. The standard defines five classes — Class 5 through Class 1 — each tied to a range of project definition completeness rather than to the effort invested. The key insight is that accuracy tracks design completion, not estimator skill alone: even the most experienced estimator cannot produce a tight number from a napkin sketch.

Without this language, a stakeholder who receives a rough feasibility figure might treat it as a firm budget — and later find the project 40% over that number. AACE classes prevent that confusion by attaching explicit accuracy ranges to every estimate before anyone signs an authorization. The class reflects design-completion percentage, not hours spent, and labeling every deliverable with its class is a professional communication standard, not bureaucracy.

Class 5 and Class 4 (Early)

Class 5 is produced when project definition is essentially zero — typically 0 to 2% complete. There are no drawings to measure from, no specifications to reference, and no equipment lists to price. Estimators rely entirely on parametric relationships: cost per square foot, cost per hospital bed, cost per megawatt. The accuracy range runs roughly -20% to -50% on the low side and +30% to +100% on the high side. That width is not a failure; it is the honest range for the information available. Class 5 answers one question: is this project in the right ballpark of feasibility?

Class 4 covers roughly 1% to 15% project definition — the feasibility stage where a concept is taking shape but detailed design has barely begun. Accuracy tightens to roughly -15% to -30% low and +20% to +50% high. Methods remain heavily analogy-based: capacity-factor scaling from similar projects, system-level assemblies, or historical unit costs. The result supports a go/no-go investment decision, not a construction budget.

• Class 5: 0–2% project definition, concept screening, roughly ±30–100% range.
• Class 4: 1–15% definition, feasibility studies, roughly ±20–50% range.
• Primary methods: parametric, capacity factors, analogy to similar completed projects.
• Purpose: decide whether a project is worth pursuing further investment.

Class 3 (Budget)

Class 3 is the estimate most commonly associated with funding authorization. It covers 10% to 40% project definition — major systems are identified, a schematic or design-development set may exist, and key equipment is named, but detailed engineering is still underway. Accuracy is commonly cited around -10% to -20% low and +10% to +30% high.

The takeoff methodology shifts here. Early classes rely almost entirely on parametric rates; Class 3 blends that with assembly-level pricing and some preliminary quantity counts where design is firm enough to measure. A structural estimator might price foundations by the cubic yard while still using square-foot factors for superstructure not yet drawn. Owners often treat this as the control baseline — the number against which construction cost performance will be tracked — which means contingency must be sized to cover the remaining definition risk.

• Class 3: 10–40% project definition, budget authorization or control baseline.
• Accuracy: roughly -10% to -20% / +10% to +30%.
• Mix of assembly-level pricing and early detailed quantity counts.
• Typically the basis for capital appropriation and project funding approval.

Class 2 and Class 1 (Detailed)

Class 2 covers roughly 30% to 75% project definition. A substantial portion of engineering is complete, drawings are available for measurement, specifications are largely written, and major equipment is quoted from current market data. Accuracy narrows to roughly -5% to -15% low and +5% to +20% high. This class serves detailed control budgets and competitive bids on design-build or fast-track projects where 100% drawings are not yet issued.

Class 1 is the definitive estimate, built on 65% to 100% project definition, with accuracy typically cited at -3% to -10% low and +3% to +15% high. Every line item is measured from complete drawings or derived from firm vendor quotes. This is what a hard-bid lump sum should be: a quantity takeoff of every trade, priced with current labor rates and subcontractor quotes, with contingency sized only for execution risk. Producing a Class 1 without complete drawings is not possible — the class itself is a certification that the underlying design is sufficiently defined to support that precision.

Matching Class to Takeoff Depth

The class framework has a direct implication for takeoff depth. Early classes demand parametric judgment — knowing what a similar project cost per unit of capacity. Detailed classes require line-item quantity takeoffs from coordinated drawings. Accuracy does not improve simply by spending more time on a Class 5 estimate; if design is 10% complete, uncertainty comes from missing definition, not from the estimator.

Stating the class on a deliverable is a professional obligation. A Class 3 budget should carry more contingency than a Class 1 bid, and an owner receiving an unlabeled number has no way to size their reserve correctly. Detailed bid estimates built on strong project definition routinely achieve 5–10% accuracy — and that accuracy is credible only because the class label makes the definition level explicit.

Questions estimators actually ask