A study of the ecology and epidemiology of kiwifruit blossom blight, believed to be caused by the bacterium Pseudomonas viridiflava, was made at the Horticulture and Food Research Institute's Kumeu Research Orchard during spring 1991. Populations of Pseudomonas species, progression of disease symptoms, and micrometeorological conditions were monitored during flowering. No correlations were found between disease progress and the climate variables measured. The rate of increase in disease incidence changed with the transition of buds to flowers. Changes in the rate of increase, logit transformed, were compared with climate variables using cross-correlation. Populations of P. viridiflava were related to the number of diseased blossoms remaining on the vine. The population of a saprophytic bacterium, P. fluorescens, increased as flowering progressed and as dead flower parts became available for colonization. Other Pseudomonas species were only present in low numbers. The population of P. viridiflava and the rate of disease progress decreased at fruit set. A comparison of simulated disease progress curves (calculated using disease model formulae) with disease progress curves from real data suggested that the epidemic was polycyclic, and observations suggested that disease was spreading to unopened buds.