Public-private partnerships (PPPs) are long-term partnerships between the public and private sectors that usually involve the private sector undertaking investment projects that traditionally have been executed (or at least financed) and owned by the public sector. This paper considers alternative approaches to value-for-money tests and discusses some of the main conceptual problems associated with these tests. It explains why comparisons between private bids and a public sector comparator are difficult and prone to significant error. It is argued that tests centered on comparisons between private sector alternatives are well focussed, less prone to measurement error than other tests, and more likely to deliver the best candidate from the group it considers. The paper also considers the PPP evidence from the United Kingdom and summarises the experience of outsourcing in the private sector.