The initial and final floor surface and food block surface actually used by termites in a small container were converted to estimates of "Living Space" and "Feeding Space" densities for 1/32 to 1½ g groups of termites. The termites were provided with blocks of blotter paper as a food supply. Except where 1½ g groups exhausted the food, the survival of termites was uniformly high (averaging >96%). The growth, block weight loss, debris production and feeding rates declined progressively as densities increased; however, these rates stabilized where 1/2, 1 and 1½ g groups had equalized feeding densities. This stabilization demonstrated that feeding space density, but not living space density, affected the termite performance. The advantage and disadvantage of using a matrix in termite bioassays is discussed.