Objective To investigate the effect of glutamine on early postoperative cognitive function in elderly patients undergoing radical surgery for colorectal cancer. Methods Sixty‑six patients undergoing radical colorectal cancer surgery were divided into a glutamine group (group A) and a control group (group C), according to the random number table method, with 33 patients in each group. Both groups received the same anesthesia methods. While the group A was administered intravenous glutamine 0.5 g/kg 24 h before and 1 h after surgery, the group C received an equivalent amount of normal saline at same time points. The incidence of perioperative neurocognitive disorder (PND) within 7 d post‑surgery was noted. Neuropsychological tests [Mini‑Mental State Examination (MMSE), Digit Symbol Test, Forward Digit Span Test, Serial Addition Test, Trail Making Test, and Backward Digit Span Test] were conducted at 3 d preoperatively (T1), 1 d postoperatively (T2), 7 d postoperatively (T3). The differences between the results obtained 3 d preoperatively and T3 were calculated. Serum concentrations of neuron‑specific enolase (NSE), interleukin (IL‑6), and S100 calcium‑binding protein β (S100β) were measured by enzyme‑linked immunosorbent assay before anesthesia induction (T0), at T1, T2, and T3. Results The group C has a higher incidence of PND within 7 d post‑surgery than group A (P<0.05). In group C, the scores of the Trail Making Test at T2 and T3 were lower than those at 3 d preoperatively (all P<0.05), and the results of the Digit Symbol Test and Serial Addition Test at T3 were lower than those at 3 d preoperatively (all P<0.05). The differences among the 3 d preoperative and T3 scores for the Digit Symbol Test, Serial Addition Test, and Trail Making Test in group A were smaller than those in group C (all P<0.05). Serum concentrations of NSE and S100β in group C were high at T1 and T2 compared to the concentration at T0 (all P<0.05). In group A, serum concentrations of IL‑6, NSE, and S100β at T1 and T2, and NSE at T3, were lower than those in group C (all P<0.05). There was no significant difference in other indexes between the two groups (all P>0.05). Conclusions Glutamine can effectively reduce the incidence of early PND in patients undergoing radical colorectal cancer surgery. However, its long‑term effects on cognitive function require further observation.