Objective To compare the impact on the results tested with arterial blood gas analysis and peripheral blood with glucose meter for critical patients in ICU in different temperature states. Methods The samples of venous biochemical test, blood arterial blood gas analysis and peripheral blood of 196 cases of critical patients in ICU were collected synchronously, and measure the D-value, correlation and bias adjustment factor of glucose blood tested with a synchronous fasting blood glucose test and venous/biochemical analyzer in different temperature states and different blood glucose groups, and the results of blood glucose test were analyzed. Results In normal temperature state, hypoglycemia simultaneous rapid intravenous glucose monitoring blood glucose level results compared with the lowest positive rate 3.31%(5/151), while the pairwise comparison showed there was both statistical signifcance between hypoglycemia group and target group (χ2=38.469), hyperglycemia group and target group (χ2=15.504) when choosing a synchronous fasting blood glucose test and intravenous blood glucose test (P<0.01). In high temperatures state, hypoglycemia simultaneous rapid intravenous glucose monitoring blood glucose level results compared with the lowest positive rate 0. There was both statistical significance between hypoglycemia group and target group (χ2=18.187), hypoglycemia group and hyperglycemia group (χ2=12.857) when choose a synchronous fasting blood glucose test and intravenous blood glucose test (P<0.01). Conclusions In high temperatures state, a synchronous fasting blood glucose test can not reflect the true value of blood glucose for critical patients.