The high cost of novel treatments is the major driver of negative or restricted reimbursement decisions by healthcare payers in many countries. Costly drugs can be subject to Market Access Agreements (MAAs), which are financial (Commercial Agreements [CAs]) or outcomes-based (Payment for Performance Agreements [P4Ps] or Coverage with Evidence Development agreements [CEDs]). Outcomes in outcomes-based MAAs are assessed through changes in surrogate endpoints (SEPs) or patient-relevant endpoints (PEPs). In May 2015, we reviewed published and grey literature on MAAs between manufacturers and large, institutionalised payers from all geographical areas, and classified the schemes into CAs, P4Ps and CEDs, as well as by therapeutic area and country. Outcomes-based MAAs were further categorized by the endpoint used. Overall, we identified 143 MAAs, 56 (39.2 %) of which were pure CAs, 53 (37.1 %) were CEDs, and 34 (23.8 %) were P4Ps. Among the CEDs, 49 were PEP CEDs and four were SEP CEDs; of the 34 P4Ps, 29 were SEP P4Ps for 30 drugs, and five were PEP P4Ps for at least six drugs; and among 87 outcomes-based MAAs (CEDs + P4Ps), PEP CEDs were the most common (56.3 %), followed by SEP P4Ps (34.1 %). The high proportion of SEPs used in P4Ps contrasts with the high proportion of PEPs used in CEDs. CEDs employ PEPs and it appears that they are used to reduce uncertainty about a drug’s clinical outcomes and/or real-life use, and thus allow payers to align a product’s value with price. We argue that P4Ps do not reduce uncertainty about real-life effectiveness and can only constitute an outcome guarantee for payers if they are based on PEPs or validated SEPs.