Disease severity, based on six parameters (time from inoculation to flecking and eruption of uredinia, uredinial density and the amo'unts of urediniospoies produced), was assessed on leaf discs cut from four cultivars of Populus spp. raised with either a long (15 h) or a short {10^ h) photoperiod. The discs were inoculated with race 4 of Melampsora medusae and subsequently incubated with either a long (15 h) or a short (10 h) photoperiod. While disease severity, based OB most parameters, was lower in discs from the continuing long photoperiod (15 h pre-and 15 h post-inoculation) than from the continuing short photoperiod (10 h pre-and 10 li postinoculation), maximum severity was developed by discs given a combination of a short pre-and a long post-inoculation photoperiod. The relative contribution of the cultivar, the pre-and the post-inoculation photoperiod and their second-and third-order interactions to variation in disease severity depended on the parameter employed to assess severity. The parameters used are elements of disease monocycles and their possible significance in modelling epidemics is discussed.