Laboratory experiments are described in which the colour response of male Kosciuscola tristis to temperature was analysed. When the temperature was changed, the colour, assessed quantitatively by means of a colour chart, varied according to time-curves approaching asymptotically a new equilibrium value. The rate of change during the early stages of accommodation is correlated with the difference between the initial and final temperatures, while the mean rate of change over the first 80 per cent, of the change is related rather to the final temperature. Other things being equal, the change from dark to blue is more rapid than the reverse change. The equilibrium colour, attained after about 18 hr at constant temperature, is related to temperature by a sigmoid curve flattening out at about 10 and 27°C. Experiments in which light intensity, background colour (black or white), relative humidity, and degree of crowding were varied at constant temperature showed that these factors are without influence on the colour of the insect. Experiments in which different parts of the intact insect, isolated halves of insects, and pieces of integument in vitro were subjected to different temperatures showed that nervous or hormonal coordination is not involved in the response, the epidermal cells acting as independent effectors.