FAOSTAT Pesticides Trade & Use

Every border crossing a crop-protection chemical makes, accounted. FAOSTAT Pesticides Trade & Use is the UN Food and Agriculture Organization's record of internationally traded pesticides: what each country bought, what it sold, and in what volume - measured annually in thousand US dollars and tonnes, per country and per product, from 1961 through 2023. Where a national customs release tells one country's story, this domain holds the whole matrix: 233 countries and territories, from major agrochemical exporters down to micro-states such as Andorra.

Two design choices make the long view usable. First, values run the entire span while quantities join from 1990 onward, when the underlying basis shifted predominantly to UN COMTRADE records - so decade-over-decade value comparisons reach back six decades, and post-1990 work gets both measures side by side. Second, the item list is deliberately two-tiered: Pesticides Total (item code 1357) carries a complete quantity series for aggregate work, while separate chemically-defined compounds - including those relevant to the Rotterdam Convention on the Prior Informed Consent Procedure - sit as their own items, so trade in a specific hazardous active is isolable rather than buried in an aggregate. FAO also imputes traded quantities for Pesticides Total between 1990 and 2020 by dividing values through sub-regional, regional or global average unit values, and fills gaps for 35 non-reporting countries with mirror statistics. In Datadory's Specialty Chemicals catalog it is the only global country-year view of agrochemical flows - scored 8/10. Get a sample of this dataset cut to your countries, items and horizon.

Two observations captured during the August 2026 research pass, exactly as they arrive:

record m49 country             item_code item                            year measure                        value
---------------------------------------------------------------------------------------------------------
6      20  Andorra             1437      2,4,5-T and its salts and esters 2007 EXPORT_VALUE_1000_US          0.177
                                                                        2007 EXPORT_QUANTITY_TONNES       0.0258
8      28  Antigua and Barbuda 1437      2,4,5-T and its salts and esters 2017 IMPORT_VALUE_1000_US          0.3498
                                                                        2017 IMPORT_QUANTITY_TONNES         0.3

# item_code 1437 = one chemically-defined compound on the Rotterdam Convention watchlist
# item_code 1357 = Pesticides Total, the aggregate with a complete quantity series from 1990
# each flow also carries its own _flag column - estimated, imputed and reported rows never blend silently

Read what the shape settles immediately. Andorra's 2007 export of item 1437 is USD 177 worth, 25.8 kilograms - real rows exist at micro-state scale, not only for the big traders, which is what makes small-market screens honest. Each observation carries all six identifying dimensions plus the figures themselves, and every flow arrives as a pair: a value column in thousand US dollars beside a quantity column in tonnes, each with its own _flag column recording whether the figure was reported, estimated or imputed. Nothing is inferred between columns, so pivoting collapses cleanly along any axis - switch the key to item_code 1357 and the same grid returns the global aggregate.

Fourteen documented fields define every record, each definition verified against live data during the August 2026 research pass:

• Identity - faostat is the internal record identifier; m49_code and country_name_en fix the reporting country or territory on the UN M49 roster.
• Product - item_code identifies the pesticide (1357 = Pesticides Total) and item names it in full.
• Time - year is the reference year of the flow, annual throughout.
• Measures - four figure columns: import_value_1000_us, import_quantity_tonnes, export_value_1000_us, export_quantity_tonnes; element codes 5622, 5610, 5922 and 5910 respectively.
• Quality - each measure pairs with its own flag column (*_flag) marking reported, estimated or externally supplied figures.

The stack forms a clean natural key - country x item x year x direction x measure - which is why joins against your own country or product masters hold up without fuzzy matching. One convention worth internalizing before you build: quantities do not exist before 1990, so any panel that spans the boundary needs an explicit rule for the early decades, and unit prices are only computable where value and quantity coexist.

Four extensions ship as additional fields on request, all derivable or adjacent but outside the fourteen-field row spine: the paired FAOSTAT Pesticides Use domain - tonnes of active ingredients applied per country and year - joined onto the same country keys so application sits beside trade; unit-value derivatives (value over quantity per country-product-year) computed on delivery so price-per-tonne panels need no re-derivation; flag-decoded columns resolving every quality flag to plain language; and mirror-statistics provenance marks identifying which rows FAO filled from partner declarations rather than the reporting country's own tapes.

Name the ones you want when you request the sample and they arrive attached to the core rows rather than left as separate folders to reconcile. A scoping pass also settles the practical questions up front: how far back your particular items extend, and which countries in your universe report versus arrive through mirrors.

• Geography: 233 countries and territories declare or appear in the matrix, and mirror statistics extend coverage to 35 non-reporting countries and territories, so micro-states show up alongside the large traders instead of dropping out of the panel.
• Temporal: annual observations from 1961 through 2023. Values run the full span; quantities begin in 1990; Pesticides Total carries a complete quantity series across the modern era. The most recent revision is dated 2025-06-30, and the domain is maintained annually - treat the latest years as drafts until revisions settle.
• Granularity: one record per country x product x year, split into four measures - import value, import quantity, export value, export quantity - each carrying its own quality flag. No aggregation hides the product or the country unless you ask for it.

One analytical caveat belongs next to the temporal claim: because quantities are imputed for Pesticides Total between 1990 and 2020 using average unit values, implied unit prices computed from those rows will be smoother than reality. Keep the flags in the frame and the artifact stays visible.

API, files, or your warehouse. Daily, weekly, or hourly.

Name the countries, items and horizon when you request the sample - one Rotterdam Convention compound across all 233 countries since 1990, or Pesticides Total for a regional bloc since 1961 - and it arrives shaped to that scope with the full field dictionary and flag semantics attached. The sample ships first either way; the ongoing feed lands on whatever cadence your models run at, already filtered to the cuts you named.

A sixty-three-year trade ledger earns its keep in five specific jobs:

• Country-level market sizing - read import value and tonnes per country and year to size national crop-protection markets without buying syndicated country reports.
• Export-momentum tracking - rank which countries gain pesticide export share, product by product, over any window; full-span values make decade-over-decade comparisons possible no other source supports at this breadth.
• Regulatory and treaty monitoring - chemically-defined compounds relevant to the Rotterdam Convention sit as distinct items, so trade in a specific hazardous active is isolable.
• Price-per-tonne panels - value beside quantity yields implied unit prices per country-product-year, raw material for margin, tariff and pass-through work.
• Supply-chain exposure screening - trace which formulation-supplying countries dominate a destination's imports before assessing concentration risk in an input chain.

Each job maps back to the specialty chemicals data hub, where this record anchors the slice's agricultural-inputs trade layer alongside production-side and regulatory alternatives.

Market researchers and consultants get national crop-protection market sizing from one consistent country-product-year grid instead of stitching national customs releases. Data scientists and ML engineers get country-year panels with flags kept as features, so models learn the difference between a declared figure and an imputed one. Policy and NGO analysts get country-year trade in Rotterdam Convention-relevant compounds against treaty commitments. Journalists, academics and students get the FAO statistical record behind any global pesticide-flow claim - the citation that survives methodological review. Investors, competitive-intelligence and sales teams round it out at relevance 2: agrochemical export momentum as an exogenous signal, and destination markets opening or closing by product.

Start from the best specialty chemicals datasets ranking to see where this sits in the pool, then read the head-to-head with ECHA CHEM for where trade volumes beat regulatory substance records.

Provenance note - published by the Food and Agriculture Organization of the United Nations under the FAOSTAT program. Post-1990 rows draw predominantly on UN COMTRADE; earlier decades rest on national country trade tapes, which is why the flag columns matter more the further back you go.

Methodology note - for Pesticides Total between 1990 and 2020, FAO imputes traded quantities by dividing values through sub-regional, regional or global average unit values, and uses mirror statistics for 35 non-reporting countries and territories. Implied unit prices computed from imputed rows are therefore constructions, not observations - reconcile value against quantity before leaning on them.

Naming note - despite the name, this domain covers trade only. Pesticide use in tonnes of active ingredients lives in a separate FAOSTAT domain; Datadory can join the two onto shared country keys when you request the sample.

Completeness note - Datadory scores this record 8/10, and the field dictionary above maps fully to the verified field set captured during the August 2026 research pass. Quantity availability before 1990 and per-item depth are confirmed at sampling rather than claimed here.

Where to go next - the rail below collects the production-side, regulatory and micro-scale alternatives in this slice, plus the vocabulary pages that make the flag system legible.