Agricultural Land, Due Diligence and Project Development

Tenure

Uganda's land tenure, and what a foreign investor can hold

Article 237(1) of the Constitution vests land in the citizens of Uganda under four tenure systems: customary, freehold, mailo and leasehold. Customary land is held under community and clan norms and is often unregistered. Freehold and mailo are registered and perpetual. Leasehold is registered for a defined term.

Article 237(2)(c) provides that non-citizens may acquire leases in land in accordance with the laws prescribed by Parliament, and expressly leaves the definition of non-citizen to Parliament. The Land Act, which appears as Chapter 236 following the Seventh Revised Edition of the Principal Laws commencing 1 July 2024, provides that a non-citizen shall not be granted a lease exceeding ninety-nine years and shall not acquire or hold mailo or freehold land.

Two points of precision. Ninety-nine years is a statutory ceiling, not an entitlement or a standard grant; actual lease terms are negotiated. And the phrase non-citizen is defined more widely than most investors expect.

Owning a company is not the same as holding land rights

It follows from the above that incorporating a Ugandan company does not by itself give you the ability to hold any tenure you like. Corporate structure and land strategy are one decision, not two, and the order matters: decide the tenure you need, then draft the articles that allow the company to hold it, then incorporate, then transact. Reversing that order is expensive.

The largest practical risk

Occupancy rights do not appear on a title search

Article 237(8) of the Constitution gives lawful and bona fide occupants of mailo, freehold and leasehold land security of occupancy. The Land Act defines a lawful occupant to include a person who entered the land with the consent of the registered owner, and a bona fide occupant to include a person who occupied and used or developed land unchallenged by the registered owner for twelve years or more, or who was settled on land by government or its agent.

The consequence is stated plainly in the Act: a lawful or bona fide occupant shall not be evicted from registered land except upon an order of eviction issued by a court, and only for non-payment of the annual nominal ground rent. That ground is exhaustively narrow. There is no statutory route to remove an occupant because you wish to farm the land. Negotiated arrangements happen in practice, but you should not plan on a legal right to one, and you should not price land as though occupancy were an administrative detail.

A second trap sits alongside it. The Land Act restricts dealings in family land without the prior consent of a spouse, and defines family land partly by reference to the norms, culture, customs, traditions or religion of the family, which no register records. A transaction entered into by a purchaser in good faith and without notice that the consent requirement was not met is void, leaving the buyer with a money claim against the seller rather than the land.

Verification

Title, boundaries and the search that has to happen first

The Ministry of Lands, Housing and Urban Development operates the Uganda National Land Information System, delivered through decentralised Ministry Zonal Offices, with a public portal for searches. The Ministry publishes a fee of UGX 10,000 for each search through the public portal, and states that a title search requires the volume and folio numbers or the county, block and plot numbers.

A search returns the registered proprietor and registered encumbrances such as mortgages, caveats and leases. That is necessary and insufficient, for the reasons above. The verification sequence that actually protects money runs like this.

• Register. Confirm the registered proprietor, the tenure, the term remaining on any lease and every registered encumbrance. Confirm the person selling is the registered proprietor or is properly authorised in writing.
• Instruction. Instruct your own advocate, independent of the seller and of any agent or introducer, and instruct a registered surveyor separately.
• Boundaries. Walk the boundary with the surveyor, the seller and the neighbours present on the same day. Boundary disputes discovered after purchase are slow and expensive.
• Occupants. Ask who lives on, cultivates, grazes on or crosses the land, and ask people other than the seller. Look for houses, graves, crops, paths, grazing and water access.
• Family land. Establish whether the land is the family's ordinary residence or the source of its sustenance, and obtain documented spousal consent where required.
• Community. Speak to the local council and to neighbouring households. A purchase the surrounding community regards as illegitimate creates operational problems for years regardless of the register.
• Restrictions. Check wetland, riverbank and lakeshore status, and check the physical planning position with the district.

Our article on verifying land title and tenure in Uganda sets out the mechanics of a search in more detail. Neither that article nor this guide substitutes for instructing a Ugandan advocate.

Site assessment

Assessing the site itself

Title tells you whether you can hold the land. Site assessment tells you whether it is worth holding. These are separate exercises and both must pass.

Water

Water is the assumption most often made on the strength of a wet-season visit. Establish the source, measure its dry-season yield or flow rather than accepting a description, and ask long-term residents what happens in a bad year. Then establish the legal position: the Ministry of Water and Environment states that a groundwater abstraction permit is required for motorised or powered pumps abstracting from boreholes, wells or springs, and that a surface water permit is required for motorised pumps or for diverting or impounding more than 400 cubic metres a day. The Water Act's free-use provision covers domestic use and irrigating a subsistence garden, which is not a commercial exemption. See on-farm water sources and water storage and farm ponds.

Soil, topography and drainage

Sample and test properly rather than judging by appearance or by what grows next door. Soil results change the year-one budget through amendment cost and can rule out a crop entirely. Assess slope, erosion risk and drainage at the same time; a field that floods in one season in four is a different asset from the one you were shown. See soil testing and fertility assessment, erosion control on slopes and flooding, waterlogging and drainage.

Access, power and infrastructure

Confirm a legal and physical right of access, and drive the route loaded in the wet season before committing. Confirm distance to the grid, connection cost and reliability, or design for solar and standby generation from the outset. Uganda's national access rate was about 60 percent in 2025, with rural access far lower, so grid availability is a question rather than an assumption.

Environment and planning

Check wetland, riverbank and lakeshore protection, which directly affects where an intake, a pump house or a field boundary may lawfully sit. Screen the project with NEMA: some projects require a project brief, some a full environmental and social impact assessment, and some neither, and an unlisted project can still be screened in. Remember that the entire country is a planning area under the Physical Planning Act, so development permission and, separately, a building permit apply to farm structures too. The detail is in guide 03.

Neighbours, labour and market

Assess the surrounding land uses, including spraying by neighbours if you plan to keep bees, and the security situation. Establish labour availability at your peak weeks rather than on average, mapped against the district's own planting and harvest calendar. And measure distance to the buyer rather than to the nearest town: a good field two hours from a bad road is a logistics problem you will pay for every week. See farm labour planning and planning farm layout.

Development

Building the project in phases

Once the site passes, the temptation is to build everything at once because mobilisation is expensive. Resist it. Phased development costs slightly more per unit and removes most of the risk of building the wrong thing.

1. Secure and stabilise

   Complete the transaction, register the interest, mark boundaries, establish access and basic security, and put in place the community relationships you will rely on for years.

2. Establish water and power

   Develop the water source and secure the permit position before anything depends on it. Install the energy solution you costed rather than the one that is quickest.

3. Build the minimum viable production unit

   One block, one house, one structure. Enough to produce and to measure, not enough to lose badly.

4. Run the pilot and measure everything

   See the next section. This is the phase most often skipped and most often regretted.

5. Correct the design

   Change layout, spacing, structure, variety or process on the evidence, while changing it is still cheap.

6. Build the handling and storage layer

   Grading, drying, storage or cold chain sized to the throughput you have proven, not to the throughput you hope for.

7. Scale in absorbable steps

   Expand in increments you could absorb losing, holding record quality constant, and stopping at each step to confirm the result held.

Procurement and contractor verification

Verify suppliers and contractors with the same discipline you apply to land. Confirm the entity exists and is registered, ask for and actually call references on comparable work, inspect previous installations, and structure payments against delivered milestones rather than in advance. For inputs, the counterfeit risk is documented: the Ministry's National Seed Policy estimates that 30 to 40 percent of seed traded is counterfeit, and a measured study using Ugandan fieldwork found retail urea at 31.8 percent nitrogen against an authentic 46 percent. See verifying seed and planting material quality and fertiliser types and quality verification.

Farm management systems and monitoring

Records are infrastructure. From the first day of the pilot, keep daily field or house records; every input purchase with date, supplier, batch and price; labour by task; all treatments and vaccinations; harvest by date, quantity and grade; every sale with buyer, price, invoice date and payment date; and rainfall or irrigation applied. A notebook transcribed weekly beats a system that never gets configured. Our note on farm record keeping and gross margin analysis covers how to turn those records into a decision.

If you will not be present, decide the reporting standard before operations start: what is reported, how often, in what format, and what independent check exists on it. Absentee ownership fails through information rather than through agronomy.

The method

Validate with a Controlled Pilot, Then Scale Decisively

A pilot is not a small version of the business. It is an experiment designed to answer the specific questions desk research could not. If you cannot state in one sentence what the pilot is testing, it is not a pilot; it is a small farm you started early.

What the pilot must measure

• Biological performance on this site: establishment, growth, yield, mortality and disease incidence, measured and dated rather than estimated at the end.
• Real input costs, from real suppliers, at real delivered quantities.
• Real labour requirement, in hours by operation, including the tasks nobody plans for.
• Losses between field and store, store and buyer, and at each grading point.
• Market response: whether the buyer accepted the product, at what grade, at what price, and how quickly they paid.
• Management load, in hours of your own or your manager's time. This is the number most often ignored and most often decisive.

Failure criteria, written before you start

Write down, in advance, the results that would cause you to stop or change course: yield below a stated threshold, mortality above a stated level, the buyer rejecting more than a stated share of output, the cash cycle exceeding a stated number of days, or management time exceeding what you can supply. Criteria written after the fact are rationalisations, not decisions.

When not to scale

Do not scale when the pilot succeeded because of unrepeatable attention; when it succeeded in an unusually good season; when the buyer took the volume as a favour; when the numbers were reconstructed rather than recorded; or when the person who made it work is not the person who will run the scaled operation.

Technical success is not commercial success

A pilot that produces an excellent crop nobody pays for on time has succeeded technically and failed commercially. Judge it on cash received against cash spent, on grade accepted rather than grade produced, and on whether the result would survive being multiplied by ten without multiplying your attention by ten. When it does survive that test, scale decisively: half-committed expansion carries the cost of growth without the benefit.

Money and exit

Project economics and planning the way out

We do not publish Ugandan yields, prices, margins or returns. We searched authoritative Ugandan and international sources for enterprise budgets and found none we would stand behind, and a plausible-looking invented figure would be the most damaging thing this guide could contain. What we can give you is the cost structure to build from your own dated quotations.

Exit planning, decided at the start

Ask, before you commit, what can be sold if you stop in year two, to whom, and what is unrecoverable. Specialised structures on leased land have poor resale. A borehole, an access road and a registered lease have better resale than a bespoke processing line. Knowing which of your assets are liquid changes how much of the budget you are willing to put into each, and it is a question best answered while you still have the option to answer it differently.

Sources and further reading

Sources for this guide

Discuss your Uganda investment project