In the 1960s the benefits of government regulation of technology were believed to outweigh any costs. But recent studies have claimed that regulation has negative effects on innovation, health and consumer choice. This case study on food colours examines such claims. EFFECTS ON HEALTH were measured by allocating a hazard rating to each colour. The negative list of 1925 removed three harmful colours which were rapidly replaced, so the benefits were short-lived. Had a proposed ban been adopted in the 1860s it would have prevented many years exposure to hazardous mineral colours. The positive list of 1957 reduced the proportion of harmful coal tar dyes from 54% of the total to 20%. Regulations brought a greater reduction in hazard levels than voluntary trade action. Delays in the introduction of a positive list created a significant hazard burden. EFFECTS ON INNOVATION were assessed from patents and discovery dates. Until the 1950s food colours were adopted from textile colours. The major period of innovation for coal tar colours was between 1856 and 1910, finishing well before regulations were made in 1957, so regulations cannot be blamed for the decline. Regulations appear to have spurred the development of at least one new coal tar dye, and many new plant colours, creating a new sector of the dye industry. EFFECTS ON CONSUMER CHOICE were assessed by case studies. Coloured milk, for example, was banned despite its popularity. Regulations have restricted choice, but have removed from the market foods that were nutritionally impoverished and poor value for money. Compositional regulations provided health protection because they reduced total exposure to colours from certain staple foods. Restricting colours to a smaller range of foods would be an effective way of coping with problems of quality and imperfect toxicological knowledge today.