Land is about rights, access and opportunity.

Rights are more than a title deed.

The Act vests land in the President on behalf of Zambia’s people and recognises customary holdings. Non-citizens have specified eligibility routes; customary tenure and leasehold are different ways of holding rights.

Tenure

The basis on which rights are held, such as customary tenure or leasehold.

See History & inclusion →

Documentation

Evidence of rights. A formal title, an incomplete title and a customary certificate are not interchangeable.

See Documents & security →

Use

What happens on land: farming, homes, conservation or other activities. Use does not identify the holder.

See Crop hectares →

Nationality and control

Who holds rights, and who controls a company holding them. A tenure category does not answer this.

See International holdings →

Current-law check still needed. Later amendments, related legislation and the facts of a particular parcel would need legal review before relying on this for a transaction. This page supplies no transaction templates.

What can we say about foreign-held land?

A record of large international agricultural deals

Deal database · 1 February 2016

Signed contracts: 24 deals / 383,784 ha. Oral agreements: 2 / 5,990 ha.

The often-cited 390,074 ha includes 300 ha in a cancelled contract. The profile’s 1.6% uses agricultural land, not arable land, as its denominator.

People receiving titles are not hectares held.

Minister Jean Kapata reported annual counts of foreign nationals who acquired land and obtained title deeds. She said most foreigners bought from individual Zambians on the open market.

To assess results, we would need project-level employment, investment, consent, displacement and compensation records. Her statement does not quantify a net benefit or harm.

Holdings were concentrated within the surveyed farm sector.

Farms holding 5–100 ha accounted for about 18% of farms and 57.8% of land in the survey-covered small and medium farm sector.

The study also finds clustering around market and transport corridors. That helps explain why local access pressure can matter even where national land totals look large; its spatial correlations do not prove causation.

Documentation and security are different questions.

The paper reports limited formal documentation, demand for title among some untitled owners, and perceived insecurity. Each measure answers a different question.

We use the weighted Table 2 values. The paper’s introduction uses different rounded figures; those would need reconciling before combining them. These measures overlap and must not be stacked into one pie.

History and policy commitments are not present-day outcomes.

The policy places the familiar 94% customary / 6% state split at independence and says conversion has since changed it. It also sets inclusion goals for allocating land.

We would need allocations by sex, age, location and area, plus evidence of retained rights. Women and youths can overlap; the targets are not additive shares of the country.

Production targets also imply hectares.

These calculations connect land to the crop targets. They do not identify available land, settle existing rights or establish how much new land could be opened.

Maize

3.85 million ha2025/26 forecast, harvested hectares (USDA, citing the Ministry)

Soya beans

2.38 million ha2023 record crop, FAOSTAT-derived, harvested hectares

Wheat

141,046 ha2023/24 harvest, Ministry figures

Fixed reference: implied hectares at each crop page’s baseline yield. The explorer below also shows alternative source rows and your yield assumption.

Irrigated land: one target, two baselines.

New irrigation needed by 2031, our arithmetic (D):
250,000 / 345,000 / 500,000 ha
from the 250k baseline (total reading) / from the 155k baseline / under the 'additional' reading

• Total vs additional: 'to reach 500,000 ha by 2031' (World Bank, PDU) vs 'an additional 500,000 ha' (MoA, May 2026).
• The 'farm blocks' (Shikabeta about 36,900 ha, Luena 186,000, Kalumwange about 159,800, Kalungwishi about 202,927) are land areas, not irrigated areas.

What would change it: An MoA or FAO AQUASTAT figure with a definition.

The baselines use different definitions and are not a time series. Wheat’s irrigated share would need checking before comparing its hectares with these totals. Inspect every irrigation record →

What would we need to establish next?

Who holds land today, and where?

We would need current, comparable parcel and beneficial-ownership records, including customary coverage. Nationality, tenure and protected-area overlays would need separate treatment.

Who benefits from investment, and who bears costs?

We would need measured jobs, output and infrastructure alongside consent, displacement, compensation and livelihood outcomes, by project and community.

Are rights secure in practice?

We would need current evidence on disputes, documentation, inheritance and women’s and young people’s control of land—not just allocation targets or title counts.

How much suitable land is available for new uses?

We would need dated definitions of agricultural, arable, cultivated and irrigated land, plus existing use rights, water access and environmental constraints. Uncultivated land is not automatically available.

What about livestock, mining, tourism and sugar?

We would need target-linked grazing, licence, protected-area and national cane-area evidence. Crop hectares cannot stand in for these different land uses.

The wider source check is a completed first pass, not blanket validation. The six reopened land-evidence sources have a separate curation record; the spatial context and crop inputs carry their own sources. Unsupported ownership percentages, nationality-based allegations, speculative returns and legal templates have not been adopted.

Public evidence can support careful questions without assigning intent or fault. Send a correction or a better source →