Tablets have been prepared by use of combinations of indomethacin, Eudragit, lactose and magnesium stearate. The processes of granulation and direct compression have been compared. The effect of adding a surface active agent to the formulation was assessed by adding sodium lauryl sulphate to a sample of the powder mix and then preparing directly compressed and granulated formulations. Two granulated samples were prepared, one with surfactant inside and the other with surfactant outside of the granule. Dissolution profiles were monitored over an 8 h period at four temperatures (range 26-43°C). Apparent zero-order rate constants were calculated over the entire 8 h process in the case of the samples without surfactant. The samples with included surfactant did not fit a single zero-order release profile and were therefore described by two apparent zero-order rate constants, one over the first 2 h and then one over the period from 3 to 8 h. The thermodynamic activation parameters were calculated for each formulation. A linear relationship was observed between the initial entropy of activation and the percentage drug release at a fixed time. It would appear that the entropie term is a dominant factor in the drug release process(es). Enthalpy-entropy and enthalpy-free energy compensation analysis was employed to test the existence of a common mechanism of dissolution from each sample. A linear enthalpy-entropy compensation plot was obtained for each sample, the enthalpy-free energy plot (which is the more rigorous test) also indicated a common mechanism, but the correlation was not as good. Compensation analysis would seem to provide a useful method of comparing mechanism of release from different formulations. © 1990.