Vegetation phenology is strongly influenced by climatic factors. Climate changes may cause phenological variations, especially in the Alps which are considered to be extremely vulnerable to global warming. The main goal of our study is to analyze European larch (Larix decidua Mill.) phenology in alpine environments and the role of the ecological factors involved, using an integrated approach based on accurate field observations and modelling techniques. We present 2 years of field-collected larch phenological data, obtained following a specifically designed observation protocol. We observed that both spring and autumn larch phenology is strongly influenced by altitude. We propose an approach for the optimization of a spring warming model (SW) and the growing season index model (GSI) consisting of a model inversion technique, based on simulated look-up tables (LUTs), that provides robust parameter estimates. The optimized models showed excellent agreement between modelled and observed data: the SW model predicts the beginning of the growing season (B GS) with a mean RMSE of 4 days, while GSI gives a prediction of the growing season length (LGS) with a RMSE of 5 days. Moreover, we showed that the original GSI parameters led to consistent errors, while the optimized ones significantly increased model accuracy. Finally, we used GSI to investigate interactions of ecological factors during springtime development and autumn senescence. We found that temperature is the most effective factor during spring recovery while photoperiod plays an important role during autumn senescence: photoperiod shows a contrasting effect with altitude decreasing its influence with increasing altitude. © 2008 ISB.