In vitro models of diabetic nephropathy that assess the role of hyperglycemia on proximal tubular cell turnover commonly compare cells in a high glucose medium (25 or 30 mM) with a low glucose medium (5 mM to 6.1 mM). Any cellular growth changes observed are usually attributed to the effect of high glucose. We hypothesize that in such experiments, glucose concentrations in the low glucose medium may decline during the course of the experiments to levels which inhibit cell growth leading to the comparative conclusion that high glucose induces hyperplasia and/or hypertrophy. In this study, primary cultures of human proximal tubular epithelial cells (PTEC) and immortalized HK-2 cells were exposed to low (5 mM) or high (17 mM, 30 mM or 47 mM) glucose for up to 6 days (PTEC) and 48 hr (HK-2). When culture media were not replenished, low glucose induced a significant increase in necrosis and release of LDH and a decrease in proliferation, metabolic activity and protein content without any changes in apoptosis. High glucose media failed to induce any of these changes. Glucose was undetectable in the low glucose culture medium after 72 hr. No significant differences were observed between any of the treatment groups when culture media were replenished daily. We conclude that regular replenishment of culture media is necessary to prevent the emergence of artifactual and misleading differences between high and low glucose groups. The current knowledge of the pathophysiology of high glucose based on cell culture systems may need to be re-evaluated.