Controversial data have been obtained with direct cellular protection by prostaglandins and sulfhydryls. In the present studies, we compared the merits and liabilities of currently available cell viability assays, some of which have not been previously employed in studies of the gastric mucosa. We also tested the hypothesis that length of incubation of isolated cells with protective agents might influence the degree of cellular damage. Gastric mucosal cells were isolated from nonfasted rats and digested with various concentrations of pronase and/or EGTA. Cell viability was assessed by trypan blue and fast green exclusion, fluorescein diacetate hydrolysis, chromium release, LDH release, mitochondrial succinate dehydrogenase activity and nuclear fluorescence induced by ethidium bromide. The experiments revealed that both pronase and EGTA are needed to obtain mucosal cells with optimal yield and initial viability and that sequential additions of pronase (30 min) and EGTA (30 min), rather than their combination (60 min), increased viability without decreasing yield. Cell viability and yield were better when pronase was used in the first incubation. For LD50 determinations, a cell suspension was incubated with ethanol (0%-15%) for 5 min. For studying the effects of protective agents, the cells were pretreated with 16,16-dmPGE2 or cysteamine HCl at 37 degrees C for 30 min before a 5-min exposure to 7.5% or 15% ethanol. The LD50 value for ethanol injury was approximately 15% for all assays except LDH, where the LD50 value was 8.5%. Preincubating gastric mucosal cells for 30 min with 16,16-dmPGE2 or cysteamine resulted in no preservation of cell viability. However, when cells were preincubated with one of the protective agents for 60 min and then exposed to 8% or 10% ethanol for 5 min, partial protection was observed when assessed by succinate dehydrogenase activity and, in certain cases, by LDH release. We conclude that all seven cell viability assays yield a measurable LD50 value for ethanol-induced cell injury, but the results may vary by as much as 82%. Low concentrations of both pronase and EGTA are needed to obtain isolated mucosal cells with both high yield and initial viability. Biochemical measures of mitochondrial activity and nuclear damage provide reliable evidence of cell viability and should be used to complement membrane permeability assays. Long preincubation of cells (60 min) with protective agents resulted in only slight protection of mitochondrial function, in contrast to the rapid induction of gastroprotection seen with these compounds in vivo. We therefore surmise that processes that contribute to organ protection occur faster and more efficiently than those that control direct cell injury and protection.