Thiazolidinediones cause sodium retention and edema by a direct effect on the kidneys. The aim of this study was to use the technique of head-out water immersion to investigate the effects of rosiglitazone on sodium and volume homeostasis in subjects with type 2 diabetes mellitus. The volume expansion response to water immersion was compared with the response on a non-immersion control day in 12 nondiabetic male subjects and 8 diet-controlled male type 2 diabetic subjects with hourly blood and urine sampling over a 4-h period. This was repeated after both groups had taken 4 mg of rosiglitazone daily for 7 days. Immersion produced a natriuresis in both groups (P < 0.001). An impairment of this natriuresis was seen in the diabetic subjects (P = 0.006). However, when rosiglitazone was taken, there was no significant difference in immersion-induced natriuresis compared with nondiabetic controls (P = 0.2). There was an immersion-induced rise in atrial natriuretic peptide (ANP) and urinary cyclic guanosine monophosphate (cGMP), in the healthy subjects (ANP P = 0.001, cGMP P = 0.043), which was not seen in the diabetic subjects (ANP P = 0.51, cGMP P = 0.74). Rosiglitazone restored the immersion-induced increase in cGMP excretion and rise of ANP in the diabetic group (ANP P = 0.048, cGMP P = 0.009). This study confirms that type 2 diabetic subjects have an impaired natriuretic response to acute volume expansion, which appears to be enhanced rather than diminished by rosiglitazone. This may be related to its effects in increasing natriuretic peptides and restoring the impaired cGMP excretion to volume expansion.