A survey was undertaken during 1996-1997 in different commercial perennial ornamentals in gardens and nurseries at Bolpur, Santiniketan and Sriniketan of Birbhum District, West Bengal, India, to study the fungal diseases of some commercial ornamentals. The leaf spot of Ficus religiosa (c.o. Alternaria sp.) is the first record. Fifteen fungal diseases have been formally described from West Bengal for the first time on these twelve ornamentals. Monthly dynamics was determined for occurrence, intensity and severity of these diseases. The diseases (referred as pathogens) that were the highest and fastest during warm and wet (rainy) months were: Alternaria alternata on Polyanthes tuberosa; Alternaria alternata and Septoria chrysanthemella on Chrysanthemum indicum; Colletotrichum gloeosporioides and Alternaria sp. on Dracaena deremensis; Alternaria sp. and Cercospora hibisci-manihotis on Hibiscus rosa-sinensis; Colletotrichum gloeosporioides on Calathea ornata; Alternaria sp. on Ficus religiosa; Cercospora jasminicola on Jasminum sambac; Lasiodiplodia theobromae and Diplocarpon rosae on Rosa multiflora. On the other hand, Cercospora sp. on Pothos scindapsus aureus; Alternaria alternata and Cercospora gerberae on Gerbera jamesonii; Glomerella cingulata on Ficus elastica; Alternaria tenuissima on Bougainvillea glabra were at their highest and fastest during hot and dry (summer to pre-rainy) months. A generalized 0-9 point scale was prepared and used to determine severity (= Percent Disease Index, PDI). Strongly predictive equations for severity in terms of intensity in all cases but one viz. Alternaria leaf spot of Dracaena deremensis in terms of occurrence were developed. Such relationships helped prior assessment of severity before the disease reaches the predicted level. Thus, although crop losses were neither determined nor sought to be predicted, a new methodology has been developed for indirect assessment in terms of severity as a direct function of yield loss in terms of occurrence or intensity but not yield or yield loss per se. These findings may help in building simple decision rules for management early in the season as soon as the disease appears in one case, and when some intensity has been achieved in all other cases. Where validated this approach may be a useful tool in plant protection, especially supervisory management and appropriate IPM.