The last few years have seen a major debate on the scale, extent, and causes of social polarisation in global cities and in Western societies in general. But the debate has often been characterised more by theoretical assertion than by empirical analysis. In particular, the concept of social polarisation has often been confused with inequality. The authors use General Household Survey data and New Earnings Survey data from 1979 to 1995 to examine the existence and extent of polarisation in London. It is argued that the evidence for polarisation is relatively weak, and that to the extent that polarisation exists it is asymmetric, with much greater growth in the size of groups at the top of the earnings distribution than at the bottom. But it is also argued that both London and Great Britain as a whole have seen a marked increase in earnings inequality over the last twenty-five years. Although most groups have improved earnings in real terms, the increase has been much greater at the top end of the earnings scale. As a result, the interquartile and interdecile earnings ratios have risen sharply. It is concluded that London has seen an increase in earnings inequality rather than growth of social polarisation.