Since 1991 the reform of the U.K.-NHS has been introduced cumulatively. Public funding of health care has been retained and the goal of the reformers is to improve the efficiency of resource allocation by creating competition on the supply side of the market. The introduction of more autonomous Trust hospitals, general practice fund holders (GPFH) and the purchaser-provider divide is described. The policy contradictions in the implementation of the reforms are analyzed: the incomplete utilization of population weighted funding, the absence of a strategy in the development of GPFHs which are at once the mavericks and the catalysts of change in the new structures, the poor articulation of pricing and contracting rules, the maintenance of planned labour and capital markets which facilitate cost control but frustrate resource reallocation, and the incomplete articulation of many market rules (e.g. about merger and exit). It seems that the rhetoric of the market has been submerged in legislation and managerial rules which increase the power of central government rather than delegating control to local providers and purchasers. The lessons of the U.K. reforms for future innovators in the design of health care systems are numerous. Would a new Hippocratic Oath requiring the delivery by professionals of knowledge based medicine be as efficient but have lower transactions costs than the creation of an internal market? Who should regulate the health care market and how? How can reform best be sequenced? Is reform of funding (competing purchasers) an essential ingredient in the reform process or will supply side reform alone be adequate? The price of knowledge is high but the cost of ignorance is greater: how can cost effectiveness data be produced and disseminated best to change behaviours? Should not all reform processes be evaluated? The British choose to avoid evaluation and as a consequence the lessons of the reforms are very difficult to quantify. Whilst the U.K.-NHS reforms created a lot of enthusiasm and energy its effects are difficult to disentangle from the simultaneous increases in funding and managerial reforms which began in 1983. There is little evidence from the U.K. or elsewhere that competition in health care produces improvements in resource allocation. There is the risk that such processes may undermine cost sontrol through erosion of single payer constraints and quality competition. The scope for improving resource allocation is considerable but competition like other unevaluated reform proposals, needs to be used with caution and recognised as a means and not an end in itself.