This paper provides some early results on a project designed to improve Aboriginal children's performance in mathematics, starting from their earliest introduction to number work. It explores the use of an enriched mathematics environment that minimally conflicts with traditional Aboriginal learning styles. The study is concerned with evaluating the effectiveness of a program intervention in remote Aboriginal schools, based on the results of pre- and post-interviews given to children at eight different schools in Western Australia at the beginning and end of 1989 and 1990. Comparison data with those for children at other schools are provided in this paper. The data derive from interviews with young children, and provide evidence on their performance in several key areas of early mathematics. Schools were categorized into three groups: White middle class; town Aboriginal and working class; and remote Aboriginal. The White middle class schools had the highest performance, followed by the town Aboriginal and White working class schools and the remote Aboriginal schools. There was a rather consistent gain in mean scores for most schools of around four points over the course of the first year so that existing differences between schools at the beginning of the year were still evident at the end of the year. At this stage it is difficult to conclude whether the intervention program has improved mathematics achievement for this group of remote Aboriginal children. There is at least no evidence of ‘progressive retardation’, which describes the current situation where Aboriginal children fall farther behind as they progress through school.