Aluminum phosphide (AlP) is a solid fumigant widely used in Iran as a grain preservative. When reacted with water or acids, AIP produces phosphine gas, a mitochondrial poison that interferes with oxidative phosphorylation and protein synthesis. Poisoning by AIP is one of the most important causes of fatal chemical toxicity in Iran. There are few studies in the medical literature addressing prognostic factors associated with AlP poisoning. In this prospective study conducted across a 14-month period commencing on 21st March 2006, we enrolled all patients admitted to the ICU of Loghman-Hakim Hospital Poison Center (Tehran, Iran) with AIP poisoning, no history of diabetes mellitus diagnosed before hospitalization, and normal body mass index. We recorded patient-specific demographic information, blood glucose level on presentation (before treatment), arterial blood gas (ABG) analysis, time elapsed between ingestion and presentation, ingested dose, duration of intensive care admission, and outcome data related to each presentation. We enrolled the group of patients who survived the intoxication as a control group and compared their blood glucose levels with those who died because of AlP poisoning. Data were analyzed by Statistical Product and Service Solutions (SPSS) software (Version 12; Chicago, Ilinois, USA) using logistic regression, Pearson correlation coefficient and Student's t-test. P values of 0.05 or less were considered as the statistical significant levels. Forty-five patients (21 women and 24 men) with acute AlP poisoning were included in the study. The mean age was 27.3 +/- 11.5 years (range: 14-62 years). Thirteen patients survived (29%) and 32 expired (71%). AlP poisoning followed deliberate ingestion in all patients. The time elapsed between ingestion and arrival at the hospital was 3.2 +/- 0.4 h. There was no significant difference between survived and non-survived groups according to age, gender, and time to treatment. However, the difference between mean blood glucose levels in survived (143.4 +/- 13.7 mg/dL) and non-survived (222.6 +/- 20 mg/dL) cases was statistically significant (P = 0.021). There was no significant correlation between blood glucose level and time to treatment, age, gender, pH, HCO3 concentration, and ingested dose. Twenty-three (71.9%) of non-survived and four (30.8%) of survived patients had a blood glucose level greater than 140 mg/dL. After adjusting according to age, gender, ingested dose, pH and HCO3 concentration The odds ratio for hyperglycemia as a risk factor for death was 5.7 (CI of 1.4-23.4). In our study, patients who succumbed to AIP poisoning had significantly higher mean blood glucose levels than those who survived. This correlation of hyperglycemic effect and mortality suggests that it may be useful in guiding risk assessment and treatment of AIP poisoning. Management of hyperglycemia may have a useful role in treatment of these patients by allowing increased entrance of glucose into cells and reducing oxygen consumption.