A CRADA — a Cooperative Research and Development Agreement — lets a company and a federal laboratory work on a project together, sharing people, facilities and results. It is a genuinely useful instrument and a frequently misunderstood one.
What a CRADA is and is not
A CRADA is a collaboration agreement. Each side contributes: the lab brings personnel, equipment and facilities; the company typically brings personnel, materials, and often funds. Under the enabling legislation the lab cannot contribute funds to the company. That is the most common misunderstanding — a CRADA is not a grant, and signing one does not put money in your account.
It is also not a procurement. You are not buying a service with a deliverable and a warranty; you are agreeing to pursue a shared research objective, and the result may not be what either side hoped. If you need a defined deliverable on a schedule, a work-for-others or technical services agreement is usually the better instrument. Ask the lab's technology transfer office which mechanism fits — that office exists to do exactly this, and our overview of technology transfer explains how it thinks.
Facilities, expertise, and IP terms
The real draw is access. National labs and federal research centers hold instruments, test ranges, materials characterization capabilities and specialist expertise that no small company could justify buying. For a hardware venture, a few weeks on equipment you cannot otherwise reach can be worth more than the cash value of a small grant.
IP terms are negotiated within limits set by statute. The general pattern: each party owns what its own employees invent, joint inventions are jointly owned, and the company can negotiate an option for an exclusive license to lab inventions arising under the agreement. The government retains rights for its own purposes. Terms differ by lab and by agency, and the negotiation takes longer than founders expect — plan for months, not weeks. Have counsel review the IP article; this is general information only, and the license terms have to fit the rest of your IP strategy.
How it differs from an STTR, and when to pursue one
An STTR is a funded award that requires formal collaboration with a research institution, with a statutory split of the work between the company and the partner. A CRADA is an unfunded collaboration instrument that stands on its own. The two can coexist, but they answer different questions, as our comparison of SBIR and STTR sets out.
Pursue a CRADA when you need capability rather than cash: a test facility, a measurement you cannot make, a scientist whose specialty is central to your problem. Skip it when you need money, when you need a firm delivery date, or when the lab's interest in publishing conflicts with your need to keep something quiet. The same calculus applies to university research partnerships, which are often easier to start.
Projects House helps companies scope lab collaborations so the engineering questions are defined before the agreement is signed. Start with the contact form.