Animal production in developed countries has changed considerably over the last 50 years. In Europe, the changes have been brought about by public policies favouring more abundant, cheaper food. As a result of these policies animal production became much more efficient, as measured by the cost of producing each egg, each kilogram of meat, each litre of milk, and so on. The pressure for efficiency has more recently become market-driven, with competition between producers and between retailers to sell food as cheaply as possible. It has thereby acquired its own momentum. In many ways, this can be viewed as a success story. Thus consumers in the developed world are able to buy animal products at prices that are low relative to those charged in the past. Efficiency has been achieved by intensification. Farms have grown in size and now keep more animals per unit area. Farming methods have been automated. Beyond automation, other features of contemporary animal farming reduce labour costs – consider, for example, cages and other types of housing designed to control the behaviour of the animals and thus make their management easier. Animals have also been bred to produce meat, milk and eggs faster and with lower feed inputs. Over the last 30 years or so public awareness of what is done to farm animals in intensive animal production has grown (Appleby, 1999). This has led to responses and initiatives by various individuals and agencies concerned with agriculture and its role in the economy and society at large. Initiatives to deal with animal welfare issues have been taken by farmers’ organisations, by professional organisations such as those representing veterinarians and animal scientists, by several of the ancillary industries involved in animal production, and by various retailers and fast food chains. People with a background in modern animal production will probably have a bias towards focusing on the average. That bias is bound to be ingrained in the thinking of someone whose professional aim is to optimise commercial income. This takes us to the main topic of this paper. Different stakeholders will take different views of the ethical issues that underpin welfare measurement, and this has a number of important implications. Let us not forget that the starting point of the growing awareness of animal welfare within animal production was public concern. If it turns out that the public, or a significant section of it, views animal welfare in a way that significantly differs from the way in which it is viewed by scientists and those working in animal production, then there are likely to be problems. These problems will have to be dealt with. Otherwise initiatives to improve animal welfare may backfire. The ethical discussion of what constitutes a good animal life must be linked to public discussion of the assessment of farm animal welfare. Concepts like ‘animal welfare’ and the ‘good animal life’ are the outcome of an ongoing process of social construction. Consequently they can be seen as results of a political struggle between different interest groups, each of which tends to promote a particular definition. Those within the pig farming and food sector who wish to play a productive role in future developments should be open to the views found among the public. They should allow concepts such as that of ‘natural living’ a prominent place in the discussion. Failure to do so may result in a social controversy like the one witnessed in relation to food biotechnology in the past decades.