This paper considers the use of passive greenhouse apparatus in field experiments investigating the biological consequences of climate change. The literature contains many accounts of such experiments claiming relevance of greenhouse treatment effects to global change scenarios. However, inadequacies in microclimate monitoring, together with incomplete understanding of greenhouse modes of action, cast doubt upon such claims. Here, treatment effects upon temperature (magnitude, range, variation, rates of change), moisture (humidity, precipitation, soil water content), light (intensity, spectral distribution), gas composition, snow cover, and wind speed are reviewed in the context of Intergovernmental Panel on Climate Change (IPCC) predictions. It is revealed that greenhouses modify each of these potentially limiting factors in a complex and interactive manner, but that the relationship between this modification and forecast conditions of climate change is poor. Interpretation of biological responses, and their extrapolation to predictive models, is thus unreliable. In order that future greenhouse experiments may overcome criticisms of artefact and lack of rigour, two amendments to methodology are proposed: (1) objective-orientated design of greenhouse apparatus (2) multiple controls addressing individual environmental factors. The importance of a priori testing of microclimate treatment effects is stressed.