Most studies of ethnic residential segregation in Australian cities have us
ed single-measure indices of dissimilarity and segregation, which access th
e degree of unevenness between two maps. Segregation is a multi-dimensional
concept, however, and in this paper we introduce an alternative way of mea
suring residential concentration which incorporates several of the key dime
nsions. The procedure is illustrated using birthplace and language-used-at-
home data for collection districts in metropolitan Sydney in 1996. Results
show no evidence of extreme spatial concentration of ethnic groups.