The design, construction, and theory of a light probe for the measurement of turbulent concentration fluctuations are given. Consideration is given to the measurement of the mean concentration, as well as the necessary corrections for noise. The limitations associated with the sample volume size, frequency response, and signal‐to‐noise ratio are discussed. Recommendations for improvement beyond the present sample volume of 5.6×10−4cc, and a frequency response of 10 000 cps are given.