Quantification and horizontal distribution of air-borne inoculum of Botrytis cinerea in a rose crop in a glasshouse of 300 m 2 was studied in 1991 and 1992. Conidia of B. cinerea were caught in spore traps consisting of an agar medium selective for B. cinerea in Petri dishes placed within the crop, at flower height 1 m above the ground. Spore catches were counted as colonies, after incubation. Lesions due to conidial infection were counted on petals of rose flowers, also after incubation. Relative humidity (RH) and temperature within the glasshouse and global radiation and windspeed outside were recorded during the experiments. The horizontal distribution of B. cinerea in a rose crop grown under glass was fairly uniform in both years. In 1991 a clear seasonal pattern in the number of colonies could not be found. In 1992 the number of colonies were high in August, September and October. The number of lesions on rose flowers showed a distinct pattern in both years. In August, September and October many lesions were counted whereas in the other months few lesions appeared. In linear regression analysis, variation in numbers of colonies (spore catches) could not be explained by environmental factors recorded during the experiments. Linear regression accounted for 76 and 63% of the variation in the number of lesions on rose flowers in 1991 and 1992, in terms of relative humidity (positively correlated), global radiation outside the glasshouse (negatively correlated) and numbers of colonies on spore traps (positively correlated). The results in the rose crop suggest that RH, global radiation and spore density in glasshouses are important variables in regulating the numbers of lesions during storage and transport. The numbers of spores in glasshouses are dependent on the production system. A glasshouse with a system resulting in wet dead tissue on the ground give higher amount of spores in the glasshouse air and through that high numbers of lesions on flowers. On roses outside the glasshouses very high numbers of lesions were counted sometimes, mostly during and after rain showers, as a result of rain-deposition of spores onto the flowers.