We develop methods for describing distributions of income growth across individuals and for comparing changes in growth distributions over time. The methods include graphical devices (‘income growth profiles’) and dominance conditions, and also summary indices, together with associated methods of estimation and inference. Taking an explicitly longitudinal perspective, our approach illuminates clearly who are the gainers and the losers, and also provides distributionally‐sensitive assessments—ones that allow the income growth for different individuals to be weighted differently. Our empirical application shows that the pattern of income growth in Britain over the period 1992–6 was less pro‐poor than that for 1998–2002, and not significantly different from the pattern for 2001–5.