This paper is organized around brief accounts of two modelling initiatives, one with an Australian and the other with a global focus, which have contributed to the policy debate about greenhouse gas abatement. The Australian study involved the combined use of an energy technology programming model, MENSA/MARKAL, with an applied general equilibrium model of the Australian economy, ORANI-F. This interfaced suite of models was used to inform the Australian government’s consultations on environmentally sustainable development. The second study involves the Global 2100 model, which also uses a mathematical programming model of energy technology, but couples it to a macro, rather than an applied general equilibrium model. The latter study is one of several, each based on a different model, used as an input to OECD deliberations on greenhouse. After giving brief accounts of the interfaces between the environmetric and economic component modules of ORANI-F\MENSA/MARKAL and of Global 2100 and a synopsis of what insights into global warming have come out of these models, I set out my ideas about how better interfaces between econometric and environmetric models can be achieved. These suggestions relate more to the ’culture’ of modelling than to technicalities. It seems that the technical (often model-specific) problems encountered at the interfaces will be more tractable if everyone involved adopts the same computer modelling language and standards of documentation. Modellers from both sides will have to discipline themselves to keep their component models down to a size that allows the interfaced system to be solved using computer resources that are routinely available.