Material Transfer Agreements: When One Is Needed to Move Material Between Institutions

A material transfer agreement is needed when tangible research material crosses a boundary between two legally distinct institutions and no sale is taking place. That is the whole test, and most of the confusion around the instrument comes from applying it to situations it was never built for. Buying material from a commercial supplier is a sale, governed by terms of sale, an invoice and, at a border, an import file. Passing part of that material to a collaborating laboratory in the next city or the next country is a transfer between institutions, and that is what an MTA governs — formalised for exactly this case as a standing agreement, with each transfer executed on a short letter ^[1].

In several African jurisdictions the question is not one of good practice at all. Kenya's access and benefit-sharing regulations make an executed material transfer agreement a precondition for moving genetic resources out of the country, with a criminal penalty attached ^[2]. South African institutions sharing human biological material for health research are legally compelled to work from a standard agreement gazetted by the Minister of Health ^[4]. Whether a given material falls inside those regimes is a question that has to be answered before anything is packed, and it is answered wrongly in both directions with roughly equal frequency.

What a material transfer agreement is, and what it is not

An MTA is a contract about a physical thing. It names a provider institution and a recipient institution, identifies the material, states the purpose the recipient may use it for, and allocates what can go wrong afterwards: onward transfer, publication, ownership of whatever the recipient makes, and liability. That is its entire job, and a well-drafted one is short.

It is not an import permit — no regulator reads it at a border. It is not ethics clearance; that is a separate document from a separate body, and the attempt in South Africa's draft revision to write research ethics committees into the agreement as parties has been criticised in the literature as a category error ^[6]. It is not a data sharing agreement, and treating it as one is the most common structural fault in current African templates: the gazetted South African instrument has been faulted for handling data thinly inside what is fundamentally a materials contract, with a separate data transfer agreement proposed as the fix ^[4][6]. And it is not a shipping document. Nothing in an MTA moves a box, clears a border, or keeps a package cold.

When one is actually needed

Write the question as a decision rather than a policy. Most transfers resolve in one line, and settling the common cases in advance means nobody improvises on the afternoon a courier is waiting.

The two failure modes are not symmetrical. An unnecessary agreement costs weeks and some goodwill. Moving material that needed one without it can be a criminal offence, can void the receiving institution's ethics approval, and can make every publication downstream contestable. Where the classification is genuinely unclear — and for synthetic compounds it often is — put the question in writing to the office that would have to defend the answer, and keep the reply in the file.

Where the agreement stops being optional

Kenya states the rule in a single sentence, and the drafting is deliberately absolute — it overrides the rest of the instrument it sits in ^[2].

Notwithstanding any provisions contained in these Regulations, no person shall transfer any genetic resources outside Kenya unless such person has executed a Material Transfer Agreement.

The surrounding machinery matters as much as the sentence. An access application under the same regulations must carry evidence of prior informed consent from interested persons and lead agencies, plus a research clearance certificate ^[2]. Conviction for an offence under them carries imprisonment of up to eighteen months, a fine of up to three hundred and fifty thousand shillings, or both ^[2]. Note also what they do not do: no schedule prescribes the form or contents of the agreement. The obligation is to have executed one, and the drafting is left to the parties — precisely where an under-resourced institution loses ground.

The international frame behind these national rules is the Nagoya Protocol, adopted at Nagoya on 29 October 2010 and in force since 12 October 2014, which requires parties to establish clear rules and procedures for prior informed consent and for mutually agreed terms ^[3]. In practice the material transfer agreement is the document in which those mutually agreed terms are recorded. That is why an MTA in an access-and-benefit-sharing country is doing constitutional work rather than administrative work: it is not paperwork attached to a shipment, it is the shipment's legal basis.

South Africa took the other available route and standardised the text itself. Institutions sharing human biological material and associated data for health research or clinical trials must use the gazetted standard agreement as their framework, a measure argued at the time as participant protection rather than trade regulation ^[4][5]. There is a sectoral precedent worth knowing: the World Health Organization's Pandemic Influenza Preparedness Framework, adopted by the Sixty-fourth World Health Assembly on 24 May 2011, carries standard material transfer agreements as annexes so that virus sharing and benefit sharing move on one pre-negotiated text ^[8]. Standardisation of that kind is the only mechanism that reliably removes negotiating asymmetry between a large recipient and a small provider.

Why the standard template may not fit the material

Most institutional templates in circulation descend from one document. The Uniform Biological Material Transfer Agreement was published in final format in March 1995 as a master agreement that an institution executes once, after which each individual transfer is completed by a short implementing letter naming the specific material ^[1]. Its architecture is worth understanding because it is the architecture your counterparty's legal office is probably assuming: the material is defined to carry its progeny and unmodified derivatives with it, use is confined to the stated research purpose, onward distribution to a third party requires the provider's written consent, the material is supplied without warranty of any kind, and the recipient carries the liability for what happens once it arrives ^[1].

The limit is in the first word of the name. That instrument is built for biological material, and its definitions assume material that reproduces: progeny, unmodified derivatives, descendants. A compound made by chemical synthesis has no progeny, and forcing it into those definitions produces clauses that mean nothing or mean too much. The complaint is not hypothetical — the critical literature on Africa's own instruments makes it about scope and fit, one analysis of South Africa's draft revision being titled around the image of forcing a square into a circle ^[6], while a broader account describes the MTA less as a legal formality than as a technology that structures the relationship between the parties signing it ^[10]. There is no continent-wide standard text for transfers of synthetic research compounds between African institutions.

The clauses that decide who carries the risk

Read for these eight things and you have read the agreement. Everything else in a typical draft is recital, notice provisions and boilerplate ^[1][4].

• Definition of the material. Does it capture only what is in the package, or everything the recipient later makes from it? This one definition decides most downstream ownership disputes.
• Permitted purpose. A narrow purpose protects the provider; an unusably narrow one means a second agreement in six months. Name the project and leave room for the work it will obviously require.
• Onward transfer. Whether the recipient may pass material to a third institution, and on what notice. Silence is not permission, and assuming otherwise is how a two-party agreement becomes a four-party problem.
• Publication and authorship. Whether the provider reviews a manuscript before submission, on what clock, and whether a review right can become a veto. Set the window in days.
• Ownership of improvements. Who owns what the recipient creates, and what rights flow back. This is where negotiating asymmetry shows most sharply, and where a resource-limited partner most often signs away more than it realised ^[7].
• Warranty and liability. Almost every template supplies material as-is and places liability on the recipient ^[1]. Confirm the receiving institution can carry that, and that its insurer has been told.
• Term, termination and disposal. What happens to unused material when the project ends — returned, destroyed or retained — and who certifies it.
• Governing law and forum. A distant foreign forum makes enforcement theoretical for a small institution. Ask for one a party can afford to reach, or arbitration seated somewhere reachable.

Signature authority is where transfers stall

An investigator cannot bind an institution, and an MTA signed by someone without delegated authority is not an agreement — it is a piece of paper that will be disowned at the first difficulty. The evidence on this is specific rather than anecdotal. A qualitative study of fourteen researchers at two Ugandan institutions found the agreement regarded as central to fairness in collaborative research, but identified protracted bureaucratic approval processes as a barrier and power dynamics in negotiation as a persistent theme, concluding that resource-limited partners need stronger frameworks behind them rather than better intentions ^[7].

The practical response is unglamorous. Establish who signs before the shipment is planned. In institutions without a technology transfer office — most of them, outside a handful of large universities — the authority sits with a registrar, a deputy vice-chancellor for research, or a legal office that sees a few of these a year and has no template of its own. Treat execution as a lead time with a real number on it, alongside courier transit and clearance, and log the date the draft went out against the date it came back: that interval is the only figure that lets you plan the next transfer honestly. Keep the delegation of authority in the file so nobody reconstructs it two years later.

The agreement does not survive the journey

An MTA describes material in the condition it left. Across a continent where ambient temperatures run high, routings are multi-leg, and a customs hold has no published duration, what arrives is not automatically what the agreement describes. The contract should therefore carry the physical facts rather than leaving them to a courier's waybill: the packaging configuration, the temperature specification the material is being held to, the intended hold time of the shipper, the point at which risk passes from provider to recipient, what constitutes acceptance, and how many working days the recipient has to reject a consignment that arrived outside specification.

The reference worth having open while drafting that clause is the World Health Organization's technical supplement on temperature-controlled transport by road and by air, which is free and treats the transport container as a qualified system with a stated hold time, monitored throughout, with excursions assessed against specification rather than argued about afterwards ^[9]. Written into an agreement, that logic becomes a short list of obligations both sides can meet.

• The provider states the condition on despatch and the specification held to, with the monitoring device inside the insulated package rather than on the outer carton ^[9].
• The shipper's hold time is stated in the agreement and sized for the worst plausible clearance on the route, not the median one.
• The recipient confirms receipt within a named number of working days and records arrival condition, whether or not anything looks wrong.
• An excursion triggers a written assessment against the material's own limits, not an argument about fault.
• The agreement says who bears the loss if the material is unusable on arrival, and whether that means replacement, refunded freight, or nothing.

The transfer file

One folder per transfer, assembled as documents appear rather than reconstructed under pressure. It answers every later question about provenance and authority, and sits alongside the batch records kept for the material itself.

• The executed agreement, with the implementing letter or annex identifying the specific material transferred ^[1].
• Evidence of signature authority at both ends — the delegation, the office, the date.
• Any statutory permit the transfer depended on, including access permits and evidence of prior informed consent where an access regime applies ^[2][3].
• The ethics clearance covering the destined work, as a separate document with its own expiry date.
• The despatch record: packaging configuration, stated hold time, and the temperature trace from inside the package ^[9].
• The arrival record: date, condition, any excursion assessment, and the acceptance or rejection notice.
• The disposal or return certificate at end of term, which closes the file rather than leaving it open indefinitely.