This paper provides evidence on the relevance of modeling the seasonal nature of electricity intensity across Chinese regions in a suitable manner with monthly data from 2003 to 2009. In contrast to previous works, this study relaxes the assumption of deterministic seasonality, allowing for time and regional variation in the Chinese economy. In doing so, unobserved-components models are used to analyze the type of seasonality stochastic or deterministic that prevails. Regional differences in the seasonal patterns and their evolution over time are also examined. Results provide new empirical evidence on the stochastic nature of electricity intensity in the majority of the provinces. In addition, we find four main effects as regards seasonal patterns: (i) Lunar New Year, (ii) Summer, (iii) Spring, and (iv) Winter effects. In the first two effects seasonality becomes positive, thus indicating that electricity intensity increases, and the last two are negative, showing improvements in the use of electricity per unit of output. However, differences are observed between northern regions and the east-south of China. In addition, once we control our estimates for temperature and prices, no significant differences are seen in the results. Conclusions from this analysis are useful for empirical modeling in the energy-economics literature, and also for designing energy policies to improve the efficiency of the use of energy resources across Chinese regions. (C) 2013 Elsevier Ltd. All rights reserved.