When migrating to React Native, teams face a fundamental choice between brownfield (keeping the existing native app as host and introducing React Native incrementally) and greenfield (building a new React Native app to replace the existing one). Brownfield preserves continuity and provides a native fallback but adds architectural complexity. Greenfield offers a cleaner foundation and better conditions for AI-assisted development but shifts risk to parity and continuity. A third hybrid approach starts greenfield with a fixed checkpoint and keeps brownfield as a tested fallback. AI coding agents now make greenfield more viable by helping port native behavior at scale, but they don't eliminate the need for clear mapping rules, parity checks, and human ownership. The right path depends on installed-user continuity, native dependency depth, legacy roadmap pressure, verification exposure, and architecture ownership.
