Prior empirical studies have employed various econometric estimation techniques to study the environmental effect of tourism demand. Prominently, these econometric modeling techniques implicitly assume that the environmental effect of tourism is symmetrical, which could sometimes be problematic. This study, therefore, utilized two econometric estimation techniques, namely, the Pesaran et al. (2001). Bounds testing approaches to the analysis of level relationships. Journal of Applied Econometrics, 16(3), 289-326) symmetric autoregressive distributed lag (ARDL) and Shin et al. (2014). Modelling asymmetric cointegration and dynamic multipliers in a nonlinear ARDL framework. In Festschrift in Honor of Peter Schmidt, pp. 281-314. New York: Springer) nonlinear ARDL (NARDL) estimation technique to disentangle the effect of tourism demand on carbon emissions in Australia. The results from the symmetric ARDL model reveal that tourism demand significantly increases carbon emissions in the long run, indicating that a 1% increase in tourism demand contributes to a 0.155% increase in carbon emissions in the long run. Contrarily, the NARDL model shows that a positive shock (an increase) in tourism demand reduces carbon emissions while a negative shock (a decrease) in tourism demand increases carbon emissions in the long run. From the NARDL estimate, a 1% increase in tourism demand is associated with a 0.220% decline in carbon emissions, while a 1% decrease in tourism demand increases carbon emissions by 0.250%. Therefore, I argue that carbon emissions depend not only on the size of tourism demand but also on the pattern - thus the increase and decline - of tourism demand. The implications of these results for policy are discussed.