Atmospheric carbon dioxide concentrations (ACDC) are critical to the global temperature control system and climate change, with ACDC levels having increased rapidly since the start of the industrial revolution in the late 1700s, and the climate having become quite unstable with significant temperature changes. For this reason, a critical evaluation of ACDC levels is crucial for determining their impact on the global climate. The purpose of this paper is to analyse the trends and volatility in ACDC levels using daily data from 1 January 1991 to 31 December 2002 collected at two observatory stations, namely Ryori, Japan and Mauna Loa, USA. The conditional variance of ACDC levels is analysed using three multivariate generalised autoregressive conditional heteroscedasticity (GARCH) models, namely CCC, VARMA-GARCH and VARMA-AGARCH. These models are able to capture the dynamics in the conditional variance and the spillover effects in the volatility of ACDC levels across the two observatory stations. (c) 2005 IMACS. Published by Elsevier B.V All rights reserved.