AIM: To compare the recovery of psychomotor function after intravenous anesthesia with remazolam or propofol compound alfentanil in patients undergoing painless gastrointestinal endoscopy. METHODS: 78 patients undergoing painless gastrointestinal endoscopy were randomly divided into group RA and group  PA. Remiazolam or propofol combined with alfentanil were given intravenously in group RA or group PA. The blood pressure, heart rate, respiratory rate and saturation of puls oxygen were recorded before procdure (T1), during checking (T2), awaking from anaesthesia (T3) and at discharging from PACU (T4). Psychomotor function, as measured by the Trieger's dot test (TDT) and digit symbol substitution test (DSST), were evaluated before anesthesia (T1), at discharging from PACU (T4), 1 h (T5) and 2 h (T6) after checking. RESULTS: From assessment of the TDT, number of dots missed (NDM), maximum distance of dots missed (MDDM) and average distance of dots missed (ADDM) at T4, T5 were significantly lower than those at T1 in two groups. The completion rates and accuracy rates of DSST at T4, T5 were significantly lower than those at T1. Results of TDT and DSST at T6 were not significantly different to those at T1. The results of NDT, MDDM and ADDM at T4, T5 in group RA were significantly lower than those in group PA. The completion rates and accuracy rates of DSST at T4, T5 in group RA increased significantly compared with group PA. Compared to group PA, the incidence of hypotension was significantly lower in group RA. There was no significant difference in the incidence of respiratory depression between the two groups. CONCLUSION: Psychomotor function was fully recovered 2 h after surgery when remazolam compound alfentanil was used for painless gastrointestinal endoscopy. Compared with propofol, psychomotor function recovery in the remazolam group was faster and there were fewer adverse effects after surgery in group RA.