Growth hormone deficiency (GHD) is defined as a serum peak GH concentration <10 ng/mL with provocation as tested by a combination of at least two separate tests. The aim of this study was to compare two standard tests, insulin and levodopa (L-dopa), with a primary focus on specificity and accuracy. Clinical data were collected retrospectively from a review of 120 children who visited the pediatric endocrine clinic at Chonnam National University Hospital for the evaluation of short stature between January 2006 and April 2014. Subjects underwent GH provocation tests with insulin and L-dopa. Blood samples were obtained at 0, 15, 30, 45, 60, 90, and 120 min after administration, and GH levels were measured. In the insulin test, serial glucose levels were also checked, closely monitoring hypoglycemia. A total of 83 children (69.2%) were diagnosed with GHD and 37 children (30.8%) were diagnosed with idiopathic short stature (ISS). Peak GH levels were achieved an average of 45 min after the administration of insulin and L-dopa for both groups. The specificity and accuracy were 78.4% and 93.6% for the insulin test and 29.7% and 79.2% for L-dopa test, respectively. In the ISS group, the cumulative frequency of a GH cutoff value of >10 ng/mL at 120 min was 75.6% after insulin stimulation compared with 35.1% after L-dopa stimulation. Considering these results, we recommend performing the insulin test first to exclude ISS and then the L-dopa test for the diagnosis of GHD. This way, ISS patients are diagnosed after a single test, thus reducing hospital days and the burden of undergoing two serial tests.