Treatment options for gestational diabetes (GDM) are limited. In order to better understand mechanisms and improve treatments, appropriate animal models of GDM are crucial. Heterozygous db mice (db/+) present with glucose intolerance, insulin resistance, and increased weight gain during, but not prior to, pregnancy. This makes them an ideal model for GDM. However, several recent studies have reported an absence of GDM phenotype in their colony. We investigated several hypotheses for why the phenotype may be absent, with the aim of re-establishing it and preventing further resources being wasted on an ineffective model. Experiments were carried out across two laboratories in two countries (New Zealand and China), and were designed to assess type of control strain, diet, presence of the misty allele, and parity as potential contributors to the lost phenotype. While hyperleptinemia and pre-pregnancy weight gain were present in all db/+mice across the four studies, we found no consistent evidence of glucose intolerance or insulin resistance during pregnancy. In conclusion, we were unable to acquire the GDM phenotype in any of our experiments, and we recommend researchers do not use the db/+ mouse as a model of GDM unless they are certain the phenotype remains in their colony.