Current approaches to identifying and describing rural livelihood strategies, and household movements between strategies over time, in developing countries are imprecise. Here we: (i) present a new statistical quantitative approach combining income and asset data to identify household activity choice variables, characterise livelihood strategy clusters, and analyse movements between strategies, and (ii) apply the approach using an environmentally-augmented three-wave household (n = 427) level panel dataset from Nepal. Combining income and asset data provides a better understanding of livelihood strategies and household movements between strategies over time than using only income or asset data. Most households changed livelihood strategy at least once over the two three-year periods. A common pathway out of poverty included an intermediate step during which households accumulate assets through farming, petty trading, and migratory work.