Irinotecan plays a crucial role in the neoadjuvant chemoradiotherapy (nCRT) of rectal cancer, but its optimal dosing is still unclear. In this study, we included 101 eligible patients with the UGT1A1*28 genotype of UGT1A1*1*1 (74.3%) and UGT1A1*1*28 (25.7%) and UGT1A1*6 genotypes of GG (63.4%), GA (32.7%), and AA (3.9%). All patients received preoperative radiotherapy (50 Gy/25 fractions) with concurrent irinotecan (UGT1A1*1*1: 80 mg/m2; UGT1A1*1*28: 65 mg/m2) and capecitabine (CapIri). SN‐38 concentrations were measured at 1.5, 24, and 49 h post‐administration. Patients were divided into four groups (Q1–Q4) based on the SN‐38 concentration. The complete‐response (CR) rate was the primary endpoint. The analysis demonstrated that the 49 h SN‐38 concentration was relatively optimal for predicting efficacy and toxicity. The Q4 group had a significantly higher CR rate than the Q1 group (p = .019), but also higher rates of adverse events (p = .009). We screened the recommended 49 h SN‐38, with a 0.5–1.0 ng/mL concentration range. We also validated the correlation between UGT1A1*6 polymorphism and SN‐38 concentration, along with the clinical efficacy of irinotecan. In conclusion, our study identified the relatively optimal timepoint and concentration range for monitoring SN38 concentrations and revealed the clinical significance of UGT1A1*6 and UGT1A1*28 polymorphisms in guiding irinotecan administration, offering meaningful insights for personalised irinotecan dosing.