The feasibility of using modified Eudragit acrylic latexes as microporous coatings for osmotic devices was investigated. Potassium chloride tablets were coated with mixtures of Eudragit RS30D and RL30D acrylic latexes that also contained a plasticizer (triethyl citrate or acetyl tributyl citrate) and a pore-forming agent (urea). A 2(5-1) fractional factorial experimental design was employed to determine the effect of five formulation variables (RS30D:RL30D polymer ratio plasticizer type, plasticizer level, urea level, and cure) on the in vitro release rate of KCl in deionized water (di water), lag time, and coat burst strength. The RS30D:RL30D polymer ratio had the greatest effect on the release rate, and both lag time and burst strength were most affected by the urea level. Statistical optimization was performed, and a coat formulation with predicted desirable in vitro performance was prepared and tested. The in vitro release rate (di water), lag time, and coat burst strength agreed well with the prediction. Dissolutions were also performed in phosphate buffered saline (PBS; pH 7.4); several formulations released markedly slower in PBS than in di water. This discrepancy was dependent on the type of plasticizer and the amount of pore former. Only those coat formulations containing acetyl tributyl citrate as the plasticizer and a 100% urea [(g urea/g polymer solids) x 100] level exhibited similar release rates in di water and PBS. The mechanism of release from these devices was primarily osmotic, whereas the release from devices coated with a formulation containing triethyl citrate and 50% urea was not dependent on the osmotic pressure difference. Devices with an osmotic release mechanism behaved similarly in vivo and in vitro.(ABSTRACT TRUNCATED AT 250 WORDS)