With the thiazolidinediones rosiglitazone and pioglitazone a novel treatment modality for type 2 diabetes has become available in many countries. As monotherapy, fasting blood glucose and glycosylated hemoglobin (HbA1c), on average, can be improved by approximately 40 mg/dl and almost 1%, respectively. In combination with other agents their efficacy is additive. Thiazolidinediones reduce insulin resistance not only in type 2 diabetes but also in non-diabetic conditions associated with insulin resistance such as obesity. The mechanism of action involves binding to the peroxisome proliferator-activated receptor (PPAR)gamma, a transcription factor that regulates the expression of specific genes especially in fat cells but also in other tissues. It is likely that thiazolidinediones primarily act in adipose tissue where PPARgamma is predominantly expressed. Thiazolidinediones have been shown to interfere with expression and release of mediators of insulin resistance originating in adipose tissue (e.g. free fatty acids, adipocytokines such as tumor necrosis factor alpha, resistin, adiponectin) in a way that results in net improvement of insulin sensitivity (i.e. in muscle and liver). Nevertheless, a direct molecular effect in skeletal muscle cannot be excluded. Interference with transcription entails a potential for side-effect risk, that cannot definitively be assessed yet. For example, the in-vitro stimulation of adipogenic differentiation may underlie the clinical observation of weight gain. Theoretically, this may turn out to be counterproductive in the long run. However, there is not sufficient evidence from humans at the moment, especially no long-term data, to allow a conclusive statement. The hepatotoxicity observed with troglitazone, on the other hand, does not seem to be PPARgamma-mediated but secondary to toxic metabolites. Based on differences in drug metabolism this problem is relatively unlikely to occur with rosiglitazone or pioglitazone. Unexplained but not unimportant is the propensity for fluid retention. In summary, with the thiazolidinediones a novel concept for the treatment of insulin resistance is available that in theory could also be used for prevention of type 2 diabetes. Long-term data are indispensable for a final risk-benefit assessment of these substances.