OBJECTIVE:A trend analysis of associations with induced abortion. METHODS:Secondary analysis of the 1973/78 cohort of the Australian Longitudinal Study of Women's Health of women responding to two or more consecutive surveys out of five (N=9,042), using generalised estimating equations. RESULTS:New abortions dropped from 7% to 2% at surveys 4 and 5. By survey 5, 16% of respondents reported abortions, only 2% of them new. Women aged in their twenties were more likely to terminate a pregnancy if they reported less-effective contraceptives (aOR2.18 CI 1.65-2.89); increased risky drinking (aOR1.65 CI 1.14-2.38); illicit drugs ≤12 months (aOR3.09 CI 2.28-4.19); or recent partner violence (aOR2.42 CI 1.61-3.64). By their thirties, women were more likely to terminate if they reported violence (aOR2.16 CI 1.31-3.56) or illicit drugs <12 months (aOR2.69 CI 1.77-4.09). Women aspiring to be fully- (OR1.58 CI 1.37-1.83) or self-employed (OR1.28 CI 1.04-1.57), with no children (OR1.41 CI 1.14-1.75) or further educated (OR 2.08 CI 1.68-2.57) were more likely to terminate than other women. CONCLUSIONS:Abortion remains strongly associated with factors affecting women's control over reproductive health such as partner violence and illicit drug use. Implications for public health: Healthcare providers should inquire about partner violence and illicit drug use among women seeking abortion, support women experiencing harm and promote effective contraception.