In recent years there has been a tremendous growth of user-generated content. In the travel, tourism and hospitality fields, travel blogs, and later online travel reviews (OTRs), have proliferated. Many researchers have taken advantage of this abundant, spontaneous and freely accessible information source and have analysed OTRs to extract insights about attractions, hotels and restaurants to improve management and marketing policies. OTRs are tourist destination image (TDI) formation agents because they show the image perceived (and transmitted) by visitors and are consulted by prospective tourists. This research proposes a conceptual and methodological framework to analyse and measure TDI from textual and paratextual elements in OTRs. The model was applied on a random sample of a million OTRs (200,000 on Things to Do, 300,000 on Hotels, and 500,000 on Restaurants) written in English, between 2013 and 2017, by visitors of the Canary Islands.