The Law Says Yes: Why Women Are Still Dying from Unsafe Abortions in South Africa

A Note Before We Begin

This piece is not about whether abortion should be legal in South Africa. That question was settled in 1996. This piece is about what happens when a country legalises a right and then systematically fails to make it accessible; and about who bears the cost of that failure in their bodies, their health, and in some cases, their lives.

Before the research for this piece began, South African social media was searched for public discourse specifically about unsafe abortion, backstreet procedures, and the gap between the CTOP Act and its implementation. What was found was general discussion about abortion stigma and occasional references to "backstreet abortions" as cautionary warnings rather than systemic critique. What was absent was any sustained public engagement with the structural failures that drive women toward unsafe providers in a country where safe, free, legal abortion has existed for nearly three decades. This platform treats that silence as data. The women dying from septic abortions in South African public hospitals are not generating a national conversation. They are generating a maternal mortality statistic.

I. The Law

The Choice on Termination of Pregnancy Act 92 of 1996 (CTOP Act) came into effect on 1 February 1997. It replaced the Abortion and Sterilisation Act of 1975; a piece of apartheid legislation that permitted abortion only in narrow circumstances approved by a panel of physicians and restricted access almost entirely to white women with private healthcare (Kaswa & Yogeswaran, 2020, South African Family Practice; Mhlanga, 2003; Jewkes & Rees, 2005, The Lancet).

The CTOP Act was a genuine rupture. It permitted termination of pregnancy on request, with no reasons required, in the first twelve weeks of gestation (Parliament of South Africa, 1996; National Department of Health Clinical Guidelines, 2021). From thirteen to twenty weeks, termination was permitted on socioeconomic grounds, in cases of rape or incest, or where the continued pregnancy posed a risk to the woman's physical or mental health, or where the foetus showed a substantial risk of severe abnormality. Beyond twenty weeks, termination was permitted only where two practitioners confirmed severe risk to the woman or foetus (Parliament of South Africa, 1996; Kaswa & Yogeswaran, 2020; Kleinsmidt, 2023, Developing World Bioethics). Crucially, the Act provided that abortion services would be free of charge at public health facilities (National Department of Health, 2021 Clinical Guidelines; Khofi, Rucell & Matandela, Frontiers in Reproductive Health, 2026).

The 2004 and 2008 amendments expanded the Act further. The 2008 amendment extended provision of first-trimester abortion to trained registered nurses; not only midwives and doctors; in an explicit attempt to expand the workforce available to deliver TOP services and reduce dependence on a small number of willing physicians (Kaswa & Yogeswaran, 2020; Favier et al., 2018, International Journal of Gynecology & Obstetrics; National Department of Health, 2021 Clinical Guidelines). By any global standard, this is progressive legislation. The CTOP Act has been cited as a model for abortion law reform across sub-Saharan Africa (Guttmacher Institute, 2025; Kaswa & Yogeswaran, 2020; Khofi et al., 2026). It is grounded in the constitutional rights to dignity, equality, privacy, and access to reproductive health services.

The law says yes. That is where the good news ends.

II. The Collapse of the 91%

Between 1997 and 2002, abortion-related maternal mortality in South Africa fell by 91% (Mhlanga, 2003; Kaswa & Yogeswaran, 2020, South African Family Practice; Jewkes & Rees, 2005, The Lancet). This is one of the most dramatic reductions in maternal mortality ever recorded following abortion legalisation anywhere in the world. Women who had previously died of sepsis and haemorrhage from illegal procedures were accessing safe, supervised care. The data was unambiguous. The law was working.

That trajectory did not hold.

More than a decade after legalisation, the Department of Health estimated that unsafe TOPs directly resulted in 23% of maternal deaths from septic miscarriages in public health facilities between 2008 and 2010 (National Department of Health Clinical Guidelines, 2021; NCCEMD, Saving Mothers, various reports). The 2014–2016 Saving Mothers report indicated unsafe TOP as an avoidable factor in 25% of maternal deaths due to miscarriage (National Department of Health Clinical Guidelines, 2021; NCCEMD, Saving Mothers 2014–2016; Hera et al., 2025). In Gauteng province, 57% of abortion-related maternal deaths were found to be the result of illegal abortions (Commission for Gender Equality, 2024, citing Gauteng provincial data).

The 91% reduction was real. It was also not sustained. What happened between 1997 and the present is not a story about the limits of legalisation. It is a story about the limits of implementation; about a state that passed a law and then declined to build the system that would make the law mean something to the women who needed it most.

III. 50% Outside the System

The single most damning statistic in the South African abortion literature is this: between 50% and 58% of all abortions in South Africa occur outside designated health facilities (National Department of Health Clinical Guidelines, 2021; Khofi et al., 2026, Frontiers in Reproductive Health; Hera, Nojoko & Stiegler et al., 2025, Annales Médico-Psychologiques; Commission for Gender Equality, 2024).

The CTOP Act has been in force for twenty-eight years. Abortion is free. No reasons are required. Trained nurses can perform first-trimester terminations. And still; more than half of all women who terminate pregnancies in South Africa do so outside the legal, regulated, supervised system. That figure is not a measure of women's preferences. It is a measure of system failure.

Fewer than 7% of public health facilities in South Africa are performing abortions (Hera et al., 2025; Khofi et al., 2026; ScienceDirect, 2024). Nearly 20% of designated TOP sites are non-operational (Jim et al., 2023, as cited in Kaswa & Yogeswaran, 2020; Kleinsmidt, 2023, Developing World Bioethics). Women typically visit at least two facilities before finding one that will see them (Kleinsmidt, 2023; Commission for Gender Equality, 2024; Spotlight NSP, December 2023). In some rural provinces, the nearest operational TOP facility may require hours of travel; travel that costs money women do not have, on days when the designated provider may or may not be present.

A 19-year-old woman in rural KwaZulu-Natal who goes to her local clinic on a day when the designated provider is absent does not get referred smoothly to an alternative. She goes home still pregnant, with a narrowing window of legal eligibility, accumulating transport costs she cannot afford and a secret she cannot share (Spotlight NSP, 2023; Commission for Gender Equality, 2024; Khofi et al., 2026). She is not a failure of the law. She is a failure of the system that was supposed to deliver it.

IV. Conscientious Objection Without Regulation

The CTOP Act makes provision for conscientious objection: a healthcare provider who objects on moral or religious grounds to performing terminations is not compelled to do so. This is a legally and ethically defensible accommodation. What the Act does not do is regulate it (Favier et al., 2018, International Journal of Gynecology & Obstetrics; Kleinsmidt, 2023, Developing World Bioethics; Lince-Deroche et al., 2014, PMC).

The consequence of unregulated conscientious objection is not individual providers declining to perform procedures they object to. The consequence is entire facilities effectively shutting down TOP services while remaining officially designated; entire management structures using conscientious objection as an administrative tool to avoid offering a service they find inconvenient; and a cascading refusal that extends well beyond the providers legally entitled to exercise it (Lince-Deroche et al., 2014, PMC; Favier et al., 2018; Khofi et al., 2026).

A 2023–2024 mixed-methods study conducted by the Sexual and Reproductive Justice Coalition (SRJC) across seven provinces, published in Frontiers in Reproductive Health in 2026 (Khofi et al., 2026), documented what this looks like in practice. Providers reported pharmacists refusing to dispense misoprostol to women with valid prescriptions. Ward staff refusing to serve tea to women recovering from TOP procedures. Nurses refusing to assist doctors in theatre during terminations. Managers routing women requesting abortions to staff they knew would turn them away. And; most significantly; providers who were willing to offer TOP services being left isolated, unsupported, and subjected to facility-based stigma and hostility from colleagues (Khofi et al., 2026; Lince-Deroche et al., 2014; Favier et al., 2018).

Section 10(1)(c) of the CTOP Act states that any person who prevents the lawful termination of pregnancy or obstructs access to a facility for the termination of a pregnancy shall be guilty of an offence and liable on conviction to a fine or imprisonment (Parliament of South Africa, 1996; Kleinsmidt, 2023, Developing World Bioethics). In twenty-eight years of the Act's operation, this provision has not produced a single recorded prosecution for obstruction of access to TOP services (Commission for Gender Equality, 2024; Favier et al., 2018). A law that criminalises obstruction but prosecutes no one for obstruction is not functioning as a law. It is functioning as a statement of intent that the state has chosen not to honour.

V. Stigma as Infrastructure

In any city in South Africa, it is easier to find an illegal abortion advertised on a lamppost than to find information about where to access a legal one.

This is not an exaggeration. Illegal abortion advertisements; telephone numbers scrawled on walls and lampposts, printed on flyers, posted in taxi ranks; are a documented feature of urban South African streetscapes in Johannesburg, Cape Town, Durban and beyond (Commission for Gender Equality, 2024; CGE Investigation into TOP in South Africa, 2021). The Commission for Gender Equality has noted with concern "a concerning absence of investigations into these operations, allowing a systemic and organised continuity with impunity" (Commission for Gender Equality, 2024). Public facilities, by contrast, do not advertise their TOP services. There is no national public communication campaign equivalent in scale to HIV testing awareness campaigns. The right exists. The information about the right does not circulate at the same scale as the stigma that makes women afraid to exercise it (Favier et al., 2018; Hera et al., 2025; Guttmacher Institute, 2025).

Research confirmed that a lack of knowledge about the South African abortion law contributed substantially to the high proportion of abortions obtained outside formal facilities; and this was nearly a decade after decriminalisation (Guttmacher Institute, 2025; Favier et al., 2018). Two decades later, the same finding recurs in the literature. The 2023–2024 SRJC study confirmed that respondents reported "there hasn't been a concerted effort in educating the public about the legality of providing abortion," nor do public facilities advertise their TOP services (Khofi et al., 2026; Favier et al., 2018; Lince-Deroche et al., 2014, PMC).

Stigma in South Africa's abortion landscape operates not only at the level of individual shame. It operates at the level of facility culture; in how staff treat women who request terminations, in the deliberate routing of women to hostile providers, in the refusal to hang a poster (Khofi et al., 2026; Lince-Deroche et al., 2014; Commission for Gender Equality, 2024). It operates as infrastructure: it shapes where women go, what they are willing to disclose, and how long they wait before seeking help. The longer a woman waits, the later her pregnancy. The later her pregnancy, the fewer legal options remain.

VI. Geography and Who Dies

The distribution of unsafe abortion in South Africa is not random. It follows the same geography as every other form of health inequality in this country: it concentrates in the places where the public health system is weakest and in the lives of the women who have the fewest alternatives.

Poor, Black, rural women. Adolescents. Immigrant women (Khofi et al., 2026; Commission for Gender Equality Investigation into TOP in SA, 2021; Spotlight NSP, 2023). These are the demographics most likely to seek unsafe procedures; not because they do not know about the CTOP Act, though many do not; but because the system that was supposed to deliver the law has not reached them in any operational sense.

The Eastern Cape has one second-trimester public abortion facility for the entire province (Spotlight NSP, December 2023). A woman in the OR Tambo district who needs a second-trimester termination; because she was raped, because her contraception failed, because she could not get to a clinic in time; faces a journey of hundreds of kilometres to access a service that the law says is her right (Commission for Gender Equality, 2024; Kaswa & Yogeswaran, 2020; Khofi et al., 2026).

Adolescents face a compounding barrier. The CTOP Act specifically provides that a person under 18 may request an abortion without parental consent; a progressive provision that reflects the reality that young women's reproductive autonomy cannot be conditioned on family approval (Parliament of South Africa, 1996; Kaswa & Yogeswaran, 2020; National Department of Health, 2021 Clinical Guidelines). Yet provider bias toward young women; the assumption that an unmarried adolescent should have the baby, the moralising refusal to provide a service without unsolicited opinion; is well-documented in qualitative research (Khofi et al., 2026; CGE Investigation, 2021; Spotlight NSP, 2023; Favier et al., 2018). A young woman who is turned away, lectured, or made to feel ashamed does not necessarily try again at another facility. She may call the number on the lamppost.

Immigrant women face a further barrier: the documented fear that accessing any public health service risks exposing their immigration status (Khofi et al., 2026; CGE Investigation, 2021; Commission for Gender Equality, 2024). The CTOP Act does not condition TOP access on citizenship or documentation. The fear of enforcement, whether grounded in reality or not, functions as an effective barrier to access regardless of what the law says.

VII. Sepsis, HIV, and the Compounding Risk

Unsafe abortion kills through two primary mechanisms: sepsis and haemorrhage (WHO, 2022; National Department of Health Clinical Guidelines, 2021; Guttmacher Institute, 2025). Sepsis; systemic infection following an unsterile procedure; is the more common cause. It is also the more insidious: it develops over days, in a woman who may not connect her worsening fever and abdominal pain to the procedure she is afraid to disclose to the hospital staff trying to treat her (Spotlight NSP, 2023; CGE, 2024; Hera et al., 2025).

South Africa's HIV burden compounds this risk in ways that the maternal mortality statistics do not fully capture. HIV-positive women who die from septic abortions are likely recorded as HIV deaths rather than abortion-related deaths (National Department of Health Clinical Guidelines, 2021; NCCEMD, Saving Mothers, various reports; Commission for Gender Equality, 2024). The true mortality burden of unsafe abortion is therefore almost certainly higher than the official figures suggest; masked by the dominant recording category of a comorbidity, absorbed into a statistic that makes the cause of death legible to funders but illegible as a policy failure.

Regional evidence from sub-Saharan Africa indicates that between 7% and 13% of maternal deaths are linked to complications arising from unsafe abortion procedures (Khofi et al., 2026, citing WHO; Guttmacher Institute, 2025; Hera et al., 2025). The WHO estimated globally that 45% of all induced abortions were unsafe and that nearly half of all unsafe abortions occurred under the least safe conditions in Africa (WHO, 2022, as cited in Hera et al., 2025; ScienceDirect, 2024). South Africa sits at a peculiar intersection of these statistics: a country that has the legal framework to drive unsafe abortion to near zero and a public health infrastructure that has not done so.

VIII. Self-Managed Abortion: Progress or Abandonment?

In the absence of a functioning public system, the abortion landscape in South Africa has not remained static. It has adapted. And the adaptation is deeply ambivalent.

Misoprostol; an antiulcer medication that induces uterine contractions and can terminate an early pregnancy effectively when used correctly; has become widely available in South Africa's informal market (Spotlight NSP, 2023; Favier et al., 2018, International Journal of Gynecology & Obstetrics; Women's Legal Centre, 2024). Respondents in the 2018 Favier et al. study noted that many backstreet providers give women black market misoprostol and are knowledgeable enough to direct them to health facilities in cases of complications; making informal abortion "increasingly safe" compared to the surgical methods that dominated the pre-1996 era (Favier et al., 2018; Spotlight NSP, 2023). Organisations including Abortion Support South Africa offer evidence-based information and support for self-managed medical abortion, and civil society organisations are filling gaps the public system has refused to close (Spotlight NSP, 2023; Commission for Gender Equality, 2024).

This is genuine harm reduction. It is also, from a policy perspective, the state's failure wearing a different coat. The legal system was supposed to provide safe abortion through trained providers in functional facilities with aftercare protocols and follow-up. Instead, South Africa has arrived at a situation where the most accessible form of safe abortion for many women is a black market pharmaceutical taken at home, guided by a non-governmental organisation, with the public health system available only as a backstop for complications (Spotlight NSP, 2023; Favier et al., 2018; Commission for Gender Equality, 2024). That is not the realisation of a reproductive right. It is the privatisation of a public obligation onto the bodies and resources of individual women and civil society organisations.

IX. What the Law Actually Required

Section 10(1) of the CTOP Act sets out the offences and penalties. Any person who prevents the lawful termination of pregnancy, or obstructs access to a facility for the termination of a pregnancy, is guilty of an offence and liable to a fine or imprisonment (Parliament of South Africa, 1996; Kleinsmidt, 2023, Developing World Bioethics). Any person who performs a termination other than as provided for in the Act is liable to imprisonment not exceeding ten years (Parliament of South Africa, 1996).

These are the law's enforcement mechanisms. They exist. They have not been used against the people who obstruct access to TOP services inside designated public health facilities. They have not been used against managers who structure rosters to ensure no willing provider is available. They have not been used against the pharmacists who refuse to dispense misoprostol. They have been used selectively against informal providers; which has the effect of criminalising the supply of unsafe abortion without reducing demand for it, pushing providers further underground and making the procedure more dangerous rather than less (Commission for Gender Equality, 2024; Guttmacher Institute, 2025; Favier et al., 2018; Khofi et al., 2026).

The Commission for Gender Equality has consistently documented these failures. An inspection visit to a hospital in the Eastern Cape found a high rate of maternal deaths, the majority stemming from unsafe abortions, with the hospital noting that "no action is taken to hold those responsible for unsafe abortion services and facilities accountable" (Commission for Gender Equality, 2024). The Commission's response to this pattern has been documentation and advocacy. The National Department of Health's response has been updated clinical guidelines. Neither has produced prosecution under Section 10. Neither has produced measurable change in the proportion of abortions occurring outside designated facilities.

X. A Right Without Infrastructure Is Not a Right

South Africa now has the CTOP Act; the Constitution's Section 27 right to reproductive healthcare; the 2008 nursing amendment; and the National Department of Health's 2021 Clinical Guidelines. It has more relevant legal and policy infrastructure for safe abortion than almost any country in sub-Saharan Africa (Kaswa & Yogeswaran, 2020; Guttmacher Institute, 2025; Khofi et al., 2026).

What it does not have is a healthcare system in which 7% of facilities performing abortions is treated as a crisis rather than a status quo. A prosecutorial culture that uses Section 10 against those who obstruct access, not only against those who provide outside designated facilities. A national communication campaign that tells women where to access a right the state has guaranteed them. A facility management framework that treats conscientious objection as an individual accommodation rather than an institutional policy. A provincial health system in which the Eastern Cape has more than one second-trimester facility for its entire population.

The law says yes. Twenty-eight years of system design, staffing decisions, management culture, prosecutorial choices, and communication priorities have said something else entirely. Between 50% and 58% of the women who terminate pregnancies in this country are living in that gap. Some of them are dying in it.

A right without infrastructure is not a right. It is a document. South Africa has the document. It always has. What it has not built, in twenty-eight years, is the system that would make it mean something to the woman who needs it.