Phenology involves the recording of recurring natural events such as the commencement of flowering or the arrival of migratory birds, and the influence on such events by edaphic and climatic factors. Analyses of phenological data have been used to examine the impacts of climate change. These studies, however, have thus far used data concentrated in the Northern Hemisphere. Indications are that long-term phenological studies, within any discipline, are limited in Australia. Datasets spanning a century or more such as are found in Europe are unlikely in Australia, given the short period of European settlement (e.g. since 1788 for Sydney, 1835 for Melbourne). Eucalypts form the focus of known Australian phytophenological studies, as they are the dominant species both in a botanical and economic sense. This study represents the first attempt to apply the Generalised Additive Model for Location, Scale and Shape (GAMLSS) technique to study a phenological data set, with the aim, in part of detecting non-linear responses to climate change (contrasting earlier stepwise regression approaches). The flowering records of four species (Eucalyptus leucoxylon, E. microcarpa, E. polyanthemos and E. tricarpa) is used here as a case study. This data set represents a long time series, by Australasian standards, using more than 30 years of monthly readings, in excess of 400 flowering and climate time points. Regardless of the cyclicity of flowering over time, this study shows that each species flowering is significantly influenced by temperature and that this effect is non-linear. Stepwise GAMLSS showed that the main temperature driver of E. leucoxylon flowering is minimum temperature (P<0.0001), maximum temperature for E. polyanthemos (P<0.0001), both minimum and maximum temperature (P<0.0001) for E. tricarpa, and mean temperature for E. microcarpa (P<0.0001). Rainfall is not a significant predictor of flowering via the stepwise GAMLSS procedure. Importantly GAMLSS also allows for the identification of lower /upper thresholds of temperature for flowering commencement /cessation; for the estimation of long and short-term non-linear effects of climate, and for the identification of lagged cyclic effects of previous flowering. Flowering intensity of all species was positively and significantly correlated with last month's flowering (P<0.0001); and with flowering 12 months earlier for E. polyanthemos and E. microcarpa. Flowering of E. polyanthemos was negatively and significantly correlated with flowering intensity 2 and 4 months prior; in the case of E. microcarpa with flowering 6 and 8 months earlier. Overall E. microcarpa and E. polyanthemos flower more intensely in response to predicted increases in mean and maximum temperature, respectively. E. leucoxylon flowers less intensely with predicted increases in minimum temperature; E. tricarpa flowers less intensely with increased maximum temperature, but more intensely with increased minimum temperature (after accounting for maximum temperature). These four species are significantly influenced by temperature and as a consequence their flowering phenology will possibly change in response to climate change. GAMLSS add credibility to the use of phenological records to detect phenological phases, local climatic impacts on flowering and possibly global climate change per se.