Recently, efforts have been made to forecast major infection periods of Botrytis blight on lily so that the need for and timing of fungicide applications can be predicted. However, little attention has been paid to the nature and source of inoculum that initiates epidemics. Better knowledge of this aspect may lead to improved disease control by elimination or reduction of the source of inoculum. Between February and May sclerotia of Botrytis elliptica were found in or on the soil, or on plant debris in fields where lilies had been grown in the previous season. Dead lily stems in organic waste piles near the production fields bore numerous sclerotia. Sclerotia producing conidiophores with conidia were found under field conditions. Sporulation of sclerotia could be induced under conditions of high relative humidity (>90%) and temperatures between an experimental range of 5–20 °C in the laboratory. During the same period numerous apothecia were found under field conditions on sclerotia associated with lily plant debris. Ascospores obtained from these apothecia produced typical B. elliptica cultures on malt agar and were pathogenic to lily plants. Crosses made by combining single ascospore cultures resulted in the production of apothecia under laboratory conditions, indicating the presence of both mating types. The occurrence of numerous sclerotia on dead lily stems producing conidia or apothecia under field conditions suggests that plant debris play a role in the epidemiology of Botrytis blight of lily. The relative importance of conidia and ascospores for the initial infection in the field is not known yet. The presence of a perfect stage can be a source of genetic variability in the fungus, and might therefore affect disease-management strategies.