Objective To establish a simple, reliable and reproducible animal model of neonatal hypoxic-ischemic encephalopathy (HIE) with similar clinical pathological process to neonates. Methods Seven days after birth, 180 Sprague-Dawley (SD) rats were randomly divided into six groups: blank control group, experimental control group and four hypoxia groups (8, 10, 12 and 14 min hypoxia groups). Those in the experimental groups were locally anesthetized with 5% lidocaine to separate their tracheas through blunt dissection, followed by tracheal clamping with vascular clamp for 8, 10, 12 and 14 min, respectively. Rats in the experimental control group were only treated with blunt dissection of trachea. No intervention was given to the blank control group. Due to significant reduction in rat survival rate after 14 min of hypoxia, no further morphological or behavioral examination was performed in this group. Rat brain tissue sections were stained with hematoxylin-eosin (HE) 12 h after modeling. Three days after modeling, the rat brain was weighted and the apoptosis of neural cells was detected with terminal deoxynucleotidyl transferase(TdT) mediated dUTP nick end labeling (TUNEL). Morris water maze was used to screen cognitive impairment in these rats at the age of two months. One-way analysis of variance was used for statistical analysis. SNK test and Dunnett 's T3 test were performed to compare homogeneous and non-homogeneous data between groups. Results Systemic cyanosis, loss of consciousness, paled body, urinary and fecal incontinence, twitching of the limbs and tail and other abnormal behavior were induced by hypoxia. Ischemic necrosis, bleeding, nucleus shrinkage in a large number of neurons and hyperchromatic nuclei were observed in the 8, 10 and 12 min hypoxia groups. Three days after modeling, brain weights of rats in the 8, 10 and 12 min hypoxia groups were lower than those of the blank control group and experimental control group [(1.16±0.07), (1.04±0.06), (0.97±0.12), (1.31±0.06) and (1.28±0.09) g, F=36.437, P<0.001]. However, the numbers of apoptotic cortical [(22.83±4.52), (30.25±3.02), (39.18±5.04), (7.96±2.24) and (8.86±2.49)/400 scope field, F=164.532, P<0.001] and hippocampal CA3 neurons of that three hypoxia groups were higher than those of the two control groups [(14.63±2.26), (20.25±3.02), (24.81±1.98), (4.75±2.66) and (6.67±1.78)/400 scope field, F=141.026, P<0.001]. At the age of two months, rats in the 8, 10 and 12 min hypoxia groups had longer escape latency [(17.99±6.48), (23.07±9.90), (38.94±32.46), (14.37±6.06) and (12.78±7.21) s, F=26.912, P<0.001] and fewer times of platform crossings than those in the control group and experimental control group [(5.00±1.41), (4.90±1.29), (3.75±1.83), (7.57±1.16) and (7.14±1.15) times, F=14.336, P<0.001]. Conclusions Pathological changes in brain tissues and behaviors of rats after modeling are in line with the characteristics of classic animal model of HIE and similar to the clinical pathology and physiology of HIE, and this could be a new, simple, reliable and reproducible animal model of HIE, being capable of controlling the duration of hypoxia accurately. Key words: Hypoxia-ischemia, brain; Disease models, animal; Trachea; Constriction