Abstract: Objective:To assess the effect of patient-controlled intravenous analgesia (PCIA) with different postoperative analgesics on prognosis after colorectal surgery. Methods: A total of 460 colorectal cancer patients (TNM Ⅰ- Ⅱ) who underwent elective surgery within January 2010to December 2012in Chongqing Cancer Hospital were randomly divided into five groups for PCIA with sufentanil, dezocine, butorphanol, morphine, and tramadol. We evaluated the analgesic efficacy, detected NK cell activity and Th 1/Th 2 ratio from peripheral blood, and observed short-term complications and long-term cancer recurrence and metastasis. Healthy volunteers served as the control group. Results:The morphine group displayed a VAS score of less than3 in the rest state and showed the longest hospital stay and the highest incidence of pruritus ( P<0. 05). The incidence of postoperative nausea and vomiting was the highest in the tramadol group ( P<0. 05). NK cell activity and Th1/Th2 ratio decreased in all experimental groups after surgery. NK activity and Th1/Th2 ratio returned to the control level 7 days after surgery in the tramadol and sulfentanil groups and 14days after surgery in the dezocine and butorphanol groups (P>0. 05), whereas those in the morphine group remained low (P<0.05). The incidences of cancer recurrence and metastasis were ranked as follows: morphine > butorphnol > dezocine>sufentanil >tramadol. Conclusion: Tramadol and sufentanil used in PCIA after colorectal surgery could facilitate the recovery of immune function and reduced the incidence of recurrence and metastasis.