Observations were made onDendrocalamus giganteusclumps in six locations in the Kandy district of Sri Lanka when 23 out of 111 clumps flowered from 1990 to 1996. Examination of fresh florets confirmed their description by Munro in 1868 (Claytonet al., 1994. In: Dassanayake MD, Fosberg FR, Clayton WD, eds.A revised handbook of the flora of Ceylon.Vol. VIII. New Delhi: Amerind Publishing, 1–458). Flowering behaviour was similar to that reported by Macmillan (1907, Annals of the Royal Botanic Gardens, Peradeniya.Vol IV: 123–129) when the species first flowered after its introduction to Sri Lanka. Giant inflorescences bearing numerous florets developed in flowering clumps. Seed set was rare. Except for two clumps that died, the others survived after flowering. Precocious flowering was seen in a 4-year old seed-raised plant. Vegetative growth was related to the seasonal rainfall, while flowering did not appear to be related to any external factor; but the development of vegetative shoots from floral primordia ofin vitrocultured inflorescences indicated the possible reversal or deviation of flowering by external factors. Flowering behaviour did not indicate mast seeding, while the relatively large proportion of clumps that flowered ruled out sporadic flowering. It was not possible to confirm that the species had a flowering cycle of 76 years as reported by Janzen (1976,Annual Review of Ecology and Systematics7: 347–391). It was more likely to be a population of cohorts with different flowering cycles.