Long-acting intraocular gas bubbles are frequently used during vitrectomy to tamponade retinal breaks. The aim of this study was to determine the effect of nitrous oxide anaesthesia on the size and effectiveness of the post-vitrectomy gas bubble. Twenty vitrectomy procedures with injection of 12% perfluoropropane (C3F8) gas were performed. For 10 of the cases routine anaesthesia with nitrous oxide was used and for 10 cases non-nitrous anaesthesia with propofol was used. The volume of the intraocular gas bubble was estimated 24 hours post-operatively using A-scan biometry. At 24 hours the gas bubble occupied a mean of 65.1% of the eye in anaesthesia with nitrous oxide and a mean of 66.1% in anaesthesia with intravenous propofol. The wide range of values of gas-fill recorded at 24 hours makes comparison of the two groups inappropriate. Several factors may account for this spread of values, but in our opinion it is the uncontrolled leakage from the sclerostomies which is the most likely. This study suggests that anaesthesia using nitrous oxide does not adversely affect the size of the C3F8 gas bubble at 24 hours post-vitrectomy when compared with anaesthesia without nitrous oxide.