Gestational diabetes mellitus (GDM) is an established risk factor for the development of overt diabetes. Since the change in diagnostic criteria for diabetes in 1997, it is unclear whether there should be any preference for fasting or post-glucose challenge blood glucose in diagnosing diabetes after GDM. The study aimed at assessing the usefulness of both diagnostic methods in women after GDM. The study enrolled 193 women with previous GDM. Women who did not have a current diagnosis of diabetes were screened for impaired fasting glucose (IFG) and for glucose intolerance with an oral 75-g glucose tolerance test. A total of 45 (23.3%) subjects declared to be already diabetic. Of the 148 non-diabetic subjects, 141 (95.3%) had normal fasting plasma glucose, whereas four (2.8%) had IFG (i.e. FPG> or =6.1 and <7.0 mmol/l) and 3 (2.5%) had FPG> or =7.0 mmol/l. Upon OGTT, among the 141 subjects with normal FPG, 6 (4.3%) were diagnosed with diabetes and 23 (16.3%) with impaired glucose tolerance (IGT); the remaining 112 (79.5%) had normal glucose tolerance. Three out of four subjects with IFG had IGT. The sensitivities of fasting criteria for diagnosis of diabetes and IFG/IGT were 14.3% (95% CI, 8.0%-37.2%) and 17.1% (95% CI, 8.6%-19.8%), respectively. The specificities were 98.6% (95% CI, 97.9%-99.7%) and 99.1% (95% CI, 96.5%-100%), respectively. The kappa for diabetes diagnosis was 0.177 (95% CI, 0.018-0.507). For women with previous GDM, the sensitivity of the new criteria based upon fasting plasma glucose is unacceptably low. In addition, the two sets of criteria are not interchangeable. Therefore, we suggest full glucose tolerance diagnostic procedures in women after GDM, including assessment of post-glucose challenge values.