Arrangements for mid‐morning breaks on wet days, snow in the playground, windy days, etc., and their connections with pupil behaviour are part of teacher folklore. But is there any statistical relationship between the incidence of disruptive behaviour and weather conditions? This paper looks at correlations between the number of disruptive incidents on any one day, as indicated by referrals to a school quiet room, and weather variables such as maximum temperature, minimum temperature and wind speed. Support is found for previous research indicating that it is not so much the particular stable weather conditions on any one day, but rather the sudden changes in weather that may be predictive of pupil disturbance. A multi‐variable regression technique is used to identify the strength and order of weather variables predictive of the number of disruptive pupil referr