An account is given of the flower of Echium plantagineum in south-eastern Australia, including stages and timing of flowering, behaviour of raindrops in the flower and aspects of floral microclimate.The concentration of nectar solutes varied with time and site, with means varying from 2 to 62% (as g sucrose/100 g solution). There was a significant negative correlation between nectar solute concentration and ambient relative humidity: the drier the air, the more concentrated the nectar.Rates of nectar secretionperflower varied with the bagging method, with long-term bagging reducing net secretion rates, possibly because of reabsorption. Rates varied with time, day and site, with a temporal pattern of change suggesting a link between rates ofphotosynthesis and secretion. Maximum nectar secretion rates in short-term bagging experiments were ca. 300 fx.g sugar/flower/hr (equivalent to > 2 mglflower 124 hr).Secretion rate was correlated with flower density. As flower density increased, secretion rate per flower decreased; rate of sugar production per unit area increased relatively more slowly than flower density. E. plantagineum could produce > 500 mg sugarlm^lday.Honeybees foraged on E. plantagineum only at ambient air temperatures above ca. lTC unless irradiance exceeded ca. 750 W m~^. Foragers collected nectar or pollen alone, or both, with the type of visit significantly correlated with nectar solute concentration. Below 35% (as g sucrose/100 g solution) most bees took pollen only; above 40%, most took nectar.Mean standing crop of nectar was generally < 100 ixg/fiower when most bees were taking nectar, but could exceed WOO ixg/ftower when bees were absent or foraging mainly for pollen. Honeybees did not always remove all nectar from flowers they probed. Reabsorption of residual nectar may augment the following day's secretion.