The suggestion that loss of labile components from dental plaque into contacting aqueous buffer during some diffusion measurements might change the plaque's diffusion behaviour, and therefore invalidate the results, has been tested using tracer clearance. (a) Diffusion of [C-14]-acetate and (H2O)-H-3 was determined before and after a 3-4 h equilibration with neutral buffer. No differences or trends were found. (b) Mean normalized [C-14]-acetate clearance curves from live plaque for measurements before and after such equilibration were identical in shape and showed no change in diffusion coefficient during clearance, as evidenced by their fits to the theoretical clearance curve. No significant difference was found between coefficients for the two periods (paired sample test). In addition, [C-14]-lactate and (H2O)-H-3 were found to permeate plaque in a miniature diaphragm diffusion cell at rates independent of whether the plaque was bathed either in saliva supernatant, plaque fluid or neutral buffer. This lent some further support to the main findings. It is concluded that diffusion through dental plaque is little affected by the bathing fluid.