Thermal comfort standards are essential to determine a good indoor climate as well as to optimize energy use inside a building. The Predicted Mean Vote (PMV) model is used to determine the indoor comfort limits in a conditioned building. However, PMV model often exaggerates thermal sensation, as it does not include adaptation undertaken by the subjects in the real environment. These adaptive effects depend upon factors like local climatic condition, ethnicity, culture, etc. In this paper the results of adaptive thermal comfort based field studies conducted in 3 naturally ventilated office buildings of cold and cloudy climate in north east India are presented. The variation in thermal sensation, thermal preference, clothing insulation, neutral temperatures and other behavioral adaptive measures undertaken by the subjects like taking hot and cold beverages, number of showers in order to feel comfortable are discussed. The comfortable thermal sensation votes are plotted on the psychometric chart and a comparison with the comfort zone prescribed by ASHRAE is made. The subjects were found to be comfortable in cooler temperature than that prescribed in the standard, and thereby a modification in the comfort zone, which reflects the adaptive action of the subjects for the region is proposed. (C) 2017 Elsevier B.V. All rights reserved.