Abstract: The fungal pathogens Leptosphaeria maculans (Desm.) Ces. et de Not. and L. biglobosa Shoemaker & Brun are responsible for Phoma stem canker – the disease is regarded as the most damaging to oilseed rape (Brassica napus L.) worldwide. In Europe, rapeseed plants are mostly infected in the autumn by ascospores produced in pseudothecia. These fruiting bodies of the perfect stage are formed on dead stems of oilseed rape plants infected in a previous growing season. It was proved that fungicide treatments against these pathogens are more effective when applied during the period of mass ascospore release, which occurs after a rain event following a full pseudothecial maturation. The prediction of the rate of L. maculans and L. biglobosa fruiting body maturation is therefore an important information for the optimisation of agrotechnical and chemical practices in cultivation of oilseed rape. The prediction of L. maculans and L. biglobosa pseudothecial maturation in Poland was based on a 9 year dataset (1998-2006), comprising biological observations of fungal development and two basic weather data: mean daily temperature and rainfall, beginning at harvest time of the previous cropping season of oilseed rape. The study concerned 100 site-years with one experiment site per year between 1998 and 2003 and the average of 35.5 sites per year in the later period. From 1998 to 2006 weather data were collected at experiment locations. Since 2006 the average distance from data collection site to a weather station was 13.3 km. The prediction model for pseudothecia maturation hypothesises that the probability of pseudothecial maturation follows a Gaussian distribution, as a function of the number of cumulated days favourable for maturation. The parameterisation of the model led to the following values: minimum daily temperature = 6.0 °C; maximum daily temperature: 29.6 °C; minimum cumulated rainfall over a 12-day period = 4.0 mm and standard deviation of the number of days favourable required for pseudothecial maturation σFD =1.9 days. The efficiency of the model was greater than 0.77, which suggests that the model can be used in a decision support system.