With signifcant increases in the demand for higher education in New Zealand since the I98Os, the question of the size of the returns to investments in education has been of signifcant interest. Despite this interest, studies have been few in number and based on aggregated rnertsures from the 1960s to 1981. This study is the first to utilise micro-level data from the I991 New Zealand Census. The study employs and compares the conventional methods of regression earnings functions, and internal rate of return analyses ofprivate and social rates of return. The results indicate that returns to education are economically signifcant, at rates that are higher for females, even when adjust-ing~for hours of work and female labour force participation differences. The returns are subject to diminishing returns, and social rates of return are 1 per cent to 3 per cent lower than private returns-a result that is consistent with those of OECD countries, but among the smaller differences by international standards.