Buildings and construction together account for almost 40 percent of energy-related carbon dioxide emissions, with housing accounting for at least 17 percent of the total. While the benefits of retrofitting housing have long been known for increasing homeowner satisfaction, reducing construction-related costs, and strengthening local economies, little attention has been given to the environmental impact of a housing improvement intervention. For housing built in informal markets, that is, housing constructed outside of regulation and government oversight, this study answers the question: how much embodied carbon can be saved by improving houses for disaster resilience? Key findings include that improving a house instead of building a new one saves two-thirds of embodied carbon of an equivalent new house, and, on average, saves 18 metric tons of carbon dioxide. For the same embodied carbon budget of one new house, more than three houses can be improved and made safer. With vertical expansion (i.e. adding a second story), improving existing housing still uses, on average, 47 percent less embodied carbon than an equivalent new house. Embodied carbon savings for a preventative upgrade are, on average, 61 percent higher than a post-disaster upgrade (without expansion). When the house is also expanded vertically, savings are still 26 percent higher if the upgrade is done preventatively.