Observations were made weekly over a period of 30 years of 208 species (trees, shrubs, herbaceous plants and geophytes) from more than 1,000 growing in a garden located 18km east of the Royal Botanic Garden Edinburgh (RBGE), Scotland (lat. 55º 56ʹN: long. 3º 09ʹW). Of these species, 27 were British native or naturalised. T he First Flowering Dates (FFD) of 67 species were without significant temperature associations with variable weather; the FFDs of the other 141 species reflected, in contrast, the net outcome of 'major' associations with late winter/spring temperatures and smaller impacts of autumn/early winter temperatures. Increases in late winter and spring temperatures advanced the onset of flowering in the current year; in contrast, increases in autumn and early winter temperatures tended to be associated with delayed flowering in the following year. With stepwise regression, penalised signal regression and thermal-time models, it was possible to identify species with 'strong' associations with both air and soil temperatures and species with 'weak' associations with either air or soil temperatures. T hermal-time models for each of 120 species, whose FFDs were associated with temperature, enabled the characterisation of (1) base temperatures, Tb(°C), at, and above which, development towards open flowers is possible; and (2) thermal constants (degree days accumulated between the start of development and the onset of flowering). Together these attributes suggested that each base temperature cohort has species with widely different degree-day requirements. B etween 1978 and 2007 mean air temperatures significantly increased by 0.080°C, 0.044°C and 0.026°C yrˉ¹ in the first, second and third quarters; soil temperatures increased by 0.060ºCyrˉ¹ in the first quarter. Over the 30-year period, the trends in flowering showed the early (February/ March) flowering species flowering c. 24 days sooner; the later flowering species (April/May) advanced by only c. 12 days.