How Many Apartments Can Fit on a Development Site?
A realistic apartment count is not a zoning calculation alone. It emerges from the buildable envelope, code, circulation, systems, efficiency, and the housing mix the market can support.
A site does not have an apartment count until it has a building.
For an owner considering a New York development site, the first question is usually framed as a number: how many apartments fit? The useful answer is that the number cannot be read directly from the lot area, a zoning map, or a broker’s buildable-square-foot estimate. Zoning may set an outer limit, but the eventual count emerges only after a building is organized within that limit. The envelope, the circulation system, the unit mix, code-required egress and accessibility, facade geometry, structural grid, mechanical distribution, and the usable depth of each floor all take a share of the allowable area.
That distinction matters before an acquisition, a land bid, a rezoning strategy, or a financing conversation. A feasibility study that reports only gross buildable area can be directionally useful, but it can also create a false certainty. Two sites with the same theoretical floor area can produce materially different apartment counts and very different economics. One may support a compact, efficient double-loaded corridor building with regular perimeter; the other may lose area to a deep floor plate, awkward lot geometry, a large core, setbacks, narrow street frontage, or an existing structure that cannot accommodate the intended layout.
The defensible thesis is simple: apartment count is an architectural feasibility result, not a zoning shortcut. The earliest test should therefore convert regulatory capacity into a credible building scheme, then test that scheme against the unit sizes, amenity expectations, operational requirements, and construction interventions the project will actually require.
Start with zoning capacity, but do not stop there
Floor area ratio, commonly called FAR, is the first broad measure of capacity. In simplified terms, it relates the permitted floor area on a zoning lot to the lot’s area. In New York’s higher-density residence districts, the Zoning Resolution sets maximum residential FARs that vary by district and by whether a project qualifies for particular affordable or senior-housing provisions. The current R6 through R12 table is the appropriate source for the number, not a third-party zoning summary.
FAR is not the same as a building footprint, height, saleable area, or apartment count. A project can have floor area that is permitted in the abstract yet difficult to place in a practical massing because of street-wall rules, setbacks, yards, courts, lot coverage, special-district provisions, or the shape and adjacency of the zoning lot. On a small or irregular parcel, the limiting condition is frequently not the arithmetic of FAR but the buildable envelope that remains after these dimensional controls are applied.
An early zoning analysis should also identify whether the site is within a special purpose district, a Mandatory Inclusionary Housing area, a flood zone, a landmark district, or another overlay that changes the path to capacity. It should confirm whether the parcel is a single zoning lot, whether development rights are being assembled or transferred, and whether any existing building, easement, restrictive declaration, or prior approval affects the calculation. These are property-specific questions. A general zoning district label is not a complete feasibility opinion.
The dwelling-unit factor is a separate constraint
Even after allowable residential floor area is established, New York may limit the number of dwelling units through a dwelling-unit factor. In many multiple-dwelling cases under Zoning Resolution Section 23-52, the maximum count is calculated by dividing the maximum permitted residential floor area by a 680-square-foot factor, with a rounding rule for fractions. The same section identifies important exceptions, including certain special density areas, qualifying senior housing, and specified conversions. Those exceptions are not a license to assume unlimited density. They change one part of the zoning calculation while building code, physical layout, financing, and operations still govern what can be delivered.
For an owner, the practical implication is that a unit mix cannot be selected after zoning is “solved.” A program built around studios and one-bedrooms can collide with a unit-factor cap even when the floor plates appear generous. Conversely, a scheme that appears to support more units can lose viability when the available floor area is spread across shallow or discontinuous plates that cannot absorb a core and corridor efficiently. The unit-factor test and the test-fit should be prepared together.
Gross floor area becomes apartments only after a building gives up its necessary area
A realistic count begins by separating gross buildable area from the portion that can become dwelling-unit interiors. Every multi-family building needs non-unit space. Depending on its size and configuration, this can include entrance and lobby areas, stairs, elevators, elevator-machine or equipment space where applicable, corridors, refuse and recycling rooms, electrical and communications rooms, domestic-water and fire-protection rooms, mechanical shafts, service areas, accessible routes, and required amenity or recreation space. Some of that area may receive particular zoning treatment; it is still real space that must be planned and constructed.
The core is often the pivotal early decision. A small building may have a compact stair-and-elevator arrangement; a taller or larger building may require additional stairs, more elevators, larger shafts, and more circulation. Those requirements are not merely code diagrams. They determine how much perimeter remains available for units, how deeply units extend from windows, where bathrooms and kitchens can stack, and whether a floor can support the target mix without creating awkward residual space.
A double-loaded corridor is usually efficient only when the floor depth, corridor location, fire separations, and facade rhythm work together. A narrow building may instead require a single-loaded arrangement. A broad plate may support multiple wings or a courtyard, but courts and light-serving openings introduce their own zoning and envelope questions. None of these outcomes is inherently preferable. The point is that a planning model needs an actual circulation diagram and plausible unit bays before it treats an apartment count as investable.
Code, access, light, and services shape the count floor by floor
The New York City Construction Codes organize the obligations that turn a mass into a legal building. The Building Code includes distinct chapters for use and occupancy, fire protection, means of egress, accessibility, interior environment, energy efficiency, structural design, mechanical systems, plumbing, and elevators. The exact provisions depend on the proposal, but the early lesson is general: a test-fit that reserves no room for compliant egress, accessible circulation, shafts, risers, and equipment is not conservative. It is incomplete.
Accessibility has a direct planning consequence. It affects clearances, door approaches, bathrooms, kitchens, routes from the public way, common areas, and portions of the dwelling inventory. It should be carried by the layout from the beginning, rather than inserted after a unit plan has been optimized around a narrower clearance diagram. Egress has a similar effect. Stair locations, travel paths, rated corridors, and exit discharge all alter the core and the shape of viable units.
The facade is equally consequential. Apartment planning depends on a repeatable relationship between facade width, window placement, unit frontage, and structural spacing. A building can have enough gross floor area but too little usable perimeter for the desired number of conventional apartments. Deep interior zones can be appropriate for kitchens, baths, storage, and circulation, but their depth and ventilation strategy must be reconciled with the applicable code and the building’s mechanical system. Window patterns, party walls, lot-line conditions, and required courts can all make nominally similar floor areas perform differently.
A target unit mix is a business assumption that must survive architectural testing
Owners often begin with an average unit size. That is a useful sensitivity tool, but it is not a layout. Dividing a net-area assumption into a gross area can produce a tempting number that ignores the fact that units do not tile a plan without loss. Corners, shafts, structural offsets, party walls, facade breaks, accessible clearances, and the need for a coherent plumbing stack all introduce friction. A building with a high share of small units also usually requires more kitchens, baths, doors, meters, and partitions per square foot, changing both coordination and cost exposure.
The right early question is therefore not “how small can the apartments be?” It is “what mix can be laid out repeatedly, sold or leased credibly, and supported by the core and envelope?” The answer should be tested in several scenarios. One scenario may prioritize a larger count and smaller units. Another may reduce the count to gain more two-bedroom inventory, larger living rooms, or better corner units. A third may test whether affordable-housing provisions or a different program alter the FAR and unit-factor relationship. The preferred count is the one that remains credible when the plan, regulatory path, and underwriting are viewed together.
Existing buildings require a different form of caution
Conversion and enlargement projects are frequently described through a simple contrast: existing structure versus new construction. In feasibility terms, the more important contrast is between a building’s inherited geometry and the geometry residential use demands. Column spacing, floor-to-floor height, slab openings, facade modules, window head heights, existing stairs, elevator locations, loading access, and the capacity of shafts all affect the unit yield. A broad former office plate can offer flexibility, but it can also place substantial floor area far from the facade. A historic envelope can preserve valuable character while limiting where new openings, stairs, or mechanical penetrations can go.
Section 23-52 is notable because it identifies specified conversions among the situations without a dwelling-unit factor. That is a zoning fact, not a conclusion about achievable count. A conversion still has to satisfy its code path and prove a workable residential layout. The Existing Building Code, the building’s legal occupancy, fire-protection strategy, vertical transportation, and the scope of alteration can all change the amount of intervention required. Before valuing a conversion opportunity on the basis of a higher theoretical count, owners should test the actual windows, structure, circulation, and systems with measured existing-condition information.
What to investigate before relying on a unit count
A disciplined feasibility sequence starts with evidence, not a rendered massing. Assemble a current survey, title and zoning-lot information, zoning map and text review, tax-lot and adjacent-lot data, available drawings of existing conditions, utility information, and a record of any approvals, restrictive declarations, or landmark status. If the property contains a building, verify dimensions in the field early enough to affect the plan.
The next step is a zoning and bulk diagram that states assumptions transparently. It should show the source of allowable floor area, the relevant unit-factor calculation or exception, yards, setbacks, courts, street-wall conditions, and any unresolved special provision. Then create at least one real test-fit on representative floors, including a core, egress, shafts, plausible structure, facade bays, and the intended unit mix. The count should be reported as a range if critical information is unresolved. Precision that depends on unverified assumptions is less useful than a well-explained range.
Finally, coordinate the architectural scheme with land-use counsel, a zoning consultant when warranted, structural and MEP engineers, code consultants, cost estimators, and the project’s finance team. These disciplines do not merely validate a design after the fact. They help distinguish a count that is physically possible from one that is approvable, financeable, and buildable. That distinction is especially important before a purchase contract, land bid, or lender presentation locks in a number.
Why a range is more useful than a premature promise
At the earliest stage, it is usually more responsible to describe apartment yield as a range than as a single definitive number. The range should be tied to stated scenarios, not used to avoid a conclusion. For example, a base case can use the zoning interpretation, an efficient but plausible core, a stated mix of studios, one-bedrooms, and two-bedrooms, and an allowance for common and service areas. A second case can show the effect of a larger share of family-sized units, a more generous lobby and amenity program, or a less efficient existing floor plate. The difference between these cases is often more valuable to an owner than a single headline count because it identifies the decisions that are actually controlling value.
This approach also makes diligence more targeted. If the range turns on whether a rear-yard condition prevents a preferred massing, the next step is a zoning confirmation. If it turns on the placement of a second stair or elevator, the next step is a code and vertical-transportation review. If it turns on window availability in an existing building, the next step is a measured facade and floor-plan survey. By contrast, a single unqualified count can conceal the fact that one unresolved condition is carrying a large portion of the assumed yield.
The range should narrow as the project advances. A preliminary massing can establish whether the site merits further work. A schematic test-fit can establish a credible unit inventory. Design development and consultant coordination can refine the count around actual shafts, structural members, equipment, and wall assemblies. The count should not be presented as final simply because a spreadsheet has more decimal places. It becomes dependable when the most consequential architectural and regulatory assumptions have been checked.
Early architectural work protects optionality
The value of an early feasibility study is not limited to identifying a maximum number of apartments. It can reveal choices that preserve optionality: whether a core can shift to create better corner units, whether a setback floor should carry larger apartments rather than a diluted repetition of a typical plan, whether a facade strategy can establish a more usable bay rhythm, or whether a modest reduction in count produces a much stronger mix. Those decisions are easiest to evaluate before a contract price, capital stack, or public commitment makes the initial count difficult to revisit.
For owners and investors, the practical discipline is to ask for the drawing behind the number. A small group of clear diagrams, including the zoning envelope, representative floor plans, a core diagram, and a schedule of assumptions, will usually reveal more than a yield figure alone. It shows where the project is robust, where it is conditional, and what must be verified next. That is the information needed to treat an apartment count as a development decision rather than an optimistic conversion of square footage.
The number worth underwriting is the number a building can actually carry
The central risk in a premature unit count is not that a calculation will be slightly wrong. It is that the project team will make a consequential acquisition, valuation, or program decision on a number that has never been translated into a building. Zoning capacity establishes the field of possibility. Architecture determines how much of that capacity can become coherent apartments, with legal egress, accessible circulation, viable light and ventilation, serviceable systems, and a marketable mix.
For a site under active consideration, the most valuable early deliverable is usually a concise feasibility study that makes its assumptions visible and shows the plan behind the count. Daniel Inocente Architecture can assist owners and development teams in turning zoning capacity into an architectural test-fit before the unit count becomes an expensive commitment.
Sources
New York City Department of Buildings, 2022 Construction Codes
FAQ
Can I estimate apartment count by dividing buildable square feet by an average unit size?
Only as an early sensitivity test. It does not account for the core, stairs, elevators, corridors, shafts, structural constraints, accessibility, facade geometry, or the unit-factor rule that may apply. A credible count requires a representative floor plan.
Does a higher FAR always mean more apartments?
No. Higher FAR increases potential floor area, but the envelope, dwelling-unit factor, required circulation, floor depth, perimeter, and desired unit mix can limit how many units a particular building can carry.
Do conversions have the same apartment-density limit as new buildings?
Not always. Section 23-52 identifies specified conversions among the cases without a dwelling-unit factor. The project still must satisfy the applicable code and demonstrate a workable residential plan within the existing structure.
When should an owner commission a test-fit?
Before relying on a unit count in a land bid, purchase contract, valuation, lender submission, or major program decision. The earlier the test-fit is prepared, the more options remain available if the site cannot support the initial assumption.
