Reducing fresh gas flow during inhalational anaesthesia results in cost savings and decreases environmental impact. We are interested in the influence of fresh gas flow on the early (induction) phase of overall fresh gas flow and vapour consumption. This stage is often excluded in studies of fresh gas flow. Data were collected from 3199 sevoflurane anaesthetics over an 11-month period in four operating theatres. We determined fresh gas flow at different stages of anaesthesia, and developed an explanatory model for the influence of the 'induction' period. Following a three-month collection of baseline data we emphasised the importance of the early phase to our department repeatedly over a two-week period. We explored the relationship between fresh gas flow and total vapour usage, and used a simple mathematical model to explore the effect of changes in the fresh gas flow and duration of the 'induction' phase. Mean fresh gas flow was 1.15 l.min À1 in the baseline period and 0.91 l.min À1 in the two months following our educational effort (p = 0.0005). In the following six months, mean fresh gas flow was 1.17 l.min À1 (p = 0.7726 compared with baseline). These results were driven by changes in both fresh gas flow and duration of the initial high-flow period. We found some correlation (R 2 = 0.85) between overall fresh gas flow and vapour consumption; a 1 l.min À1 increase in fresh gas flow consumes an additional 18 ml.hr À1 of liquid sevoflurane. This preliminary study demonstrates that an episode of high fresh gas flow at the start of anaesthesia has a large and modifiable effect on overall fresh gas flow and vapour consumption. We also confirmed the linear relationship between fresh gas flow and vapour usage.