The in vitro secretion of 131I from prelabeled mouse thyroids was measured under conditions of TSH and cylic nucleotide stimulation as well as inhibition by a variety of agents. TSH, 3′,5′-cyclic AMP and dibutryl cyclic AMP all caused a dose dependent release of nonprotein 131I superimposed on a constant basal amount of 19S iodoprotein, confirming the usefulness of the Munro assay. Release of iodoprotein has been reduced to 2–6% of thyroidal 131I over a 6 hr period of incubation with good histologic preservation. Fluoride at concentrations above 3 HIM caused release of iodoprotein and blocked the response to both TSH and cyclic AMP. One mM theophylline potentiated TSH and cyclic AMP induced release but higher concentrations released iodoprotein and 10 mM theophylline blocked the physiological response to TSH. These multiple actions help explain discrepancies in previous experimental findings with fluoride and theophylline and suggest caution in their use. Chlorpromazine inhibited nonprotein 131I release by both TSH and dibutryl cyclic AMP but at higher concentrations released iodoprotein. Similar biphasic response curves were produced by the detergents sodium dodecyl sulfate and thymol. In contrast, cationic detergents released iodoproteins at lower concentrations than those at which inhibition of nonprotein I31I release occurred, and polylysine released iodoprotein without interfering with TSH stimulation of nonprotein 131I. It is concluded that thyroid membranes exposed to membrane stabilizers undergo typical stabilization-destabilization patterns characteristic of these agents. While all cellular and intercellular membranes are affected, the evidence suggests that inhibition of thyroid secretion occurs as a result of stabilization of the apical cell membrane with prevention of colloid droplet formation. (Endocrinology88: 206,1971)