Objective: To observe the effects of transient exposure to high glucose on biological behaviors of human dermal microvascular endothelial cells cultured in vitro. Methods: The dividing method and treatment of cells for the detection of all indexes in this study were as follows. Human dermal microvascular endothelial cells of the 4th passage were divided into 3 groups according to the random number table, with 12 wells in each group. Cells in control group (C) were cultured with complete culture solution containing 5 mmol/L D-glucose for 7 d. Cells in transient high glucose group (THG) were cultured with complete culture solution containing 30 mmol/L D-glucose for 2 d and complete culture solution containing 5 mmol/L D-glucose for 5 d. Cells in prolonged high glucose group (PHG) were cultured with complete culture solution containing 30 mmol/L D-glucose for 7 d. (1) The cell morphology in groups C and PHG on culture day 7 and that in group THG on culture day 2 and 7 was observed by inverted optical microscope. (2) On culture day 0, 2, 4, and 7, cell proliferation rate was determined by cell viability analyzing counter. (3) After culture day 2, the scratch experiment was performed, and the cells were further cultured. At post scratch hour (PSH) 0, 24, 48, 72, 96, and 120, the scratch area was measured, and the cell migration rates of the latter 5 time points were calculated. (4) On culture day 0, 2, 4, and 7, the cell apoptosis rate was determined by cell analyzer. (5) Cells were seeded into Matrigel to culture for 24 h after culture day 7. The formation of vessel-like structure was observed by inverted optical microscope. The length and number of branch point of vessel-like structure were calculated. (6) On culture day 2, 4, and 7, mRNA expression of vascularization-related gene tissue inhibitor of matrix metalloproteinase-3 (TIMP-3) was determined with real-time fluorescent quantitative reverse transcription polymerase chain reaction. Data were processed with analysis of variance of factorial design, analysis of variance for repeated measurement, one-way analysis of variance, and LSD test. Results: (1) Cells in group C exhibited ovary shape in cobble stone order on culture day 7. Cells in group THG exhibited long ovary shape and lost cobble stone order on culture day 2 and kept the same changes on culture day 7. Cells in group PHG exhibited long ovary shape and lost cobble stone order on culture day 7. (2) On culture day 0, there was no significant difference in cell proliferation rate among the 3 groups (F=0.23, P>0.05). On culture day 2, cell proliferation rates in groups THG and PHG were similar (P>0.05), which were significantly lower than the cell proliferation rate in group C (with P values below 0.01). On culture day 4 and 7, the cell proliferation rates in groups THG and C were similar (with P values above 0.05), which were significantly higher than those in group PHG (with P values below 0.01). (3) At PSH 24-120, the cell migration rates in groups THG and PHG were similar (with P values above 0.05), which were significantly lower than those in group C (with P values below 0.01). (4) On culture day 0, there was no statistically significant difference in cell apoptosis rate among the 3 groups (F=0.78, P>0.05). On culture day 2, cell apoptosis rates in groups THG and PHG were similar (P>0.05), which were significantly higher than the cell apoptosis rate in group C (with P values below 0.01). On culture day 4 and 7, the cell apoptosis rates in groups THG and C were similar (with P values above 0.05), which were significantly lower than those in group PHG (with P values below 0.01). (5) The length of vessel-like structure of cells in group THG was (1.84±0.10)×10⁵ μm, close to (1.82±0.11)×10⁵ μm in group PHG (P>0.05), both significantly shorter than (2.75±0.23)×10⁵ μm in group C (with P values below 0.01). The numbers of branch point of vessel-like structure of cells in groups THG and PHG were 43±5 and 46±8 respectively, which were close to each other (P>0.05) and both significantly less than 103±21 in group C (with P values below 0.01). (6) On culture day 2, 4, and 7, mRNA expressions of TIMP-3 of cells in groups THG and PHG were similar (with P values above 0.05), which were significantly lower than those in group C (with P values below 0.05). Conclusions Transient exposure to high glucose can cause metabolic memory of morphology, migration, and angiogenesis in human dermal microvascular endothelial cells cultured in vitro, resulting in sustained changes in biological behaviors. The mechanism may be related to the changes of vascularization-related genes.