Ask scientists and technicians to describe what they do in laboratories and you are reminded that science is work – it is repeated movements, lifting and carrying, painstaking and fiddly procedures, planning-out, record keeping, finding paper for the printer. We think we know what laboratories look like; there is a standard iconography: clean surfaces, glassware, instruments buzzing and flashing. Talking to scientists at length about their work in laboratories, with (in some cases) linked videos, allows us to appreciate the non-standard material culture of the laboratory: the scribbled jokes, postcards, trinkets, apparatus made of Meccano and caretakers growing flowers. Laboratory work is varied, but it is possible to identify at least three kinds of work. 1. New things (such as materials and computers) are made and tested. 2. Models (‘analogues’)are made of parts of the world, such as a tank of Lyle’s Golden Syrup to simulate the Earth’s interior, or crumpled kitchen foil to act like the rock under a glacier. 3. Natural and manufactured materials (metals, cores of ice, shells and so on) are observed and measured, often with a numerical output – a series of meaningful numbers, a trace or plot on a graph.