Certain quantities of what was referred to by South African CW scientists as "BZ," a potent psychoincapacitant, were produced at the principal South African CW research and production facility, Delta G Scientific, under the auspices of Project Coast. Some of this BZ was later tested at the main BW research, testing, and production facility, Roodeplaat Research Laboratories (RRL). However, controversial claims that this military-grade incapacitant may have been used in a January 1992 attack on Frente de Libertação de Moçambique (FRELIMO) troops near Ngungwe have not been confirmed.

In 1989 the South African Defence Force (SADF) awarded Delta G a three-year contract to produce BZ analogs "with optimal psychotomimetic properties." These were specifically developed for use as potential incapacitants that would quickly attack the central nervous system and thereby disorient targeted persons via hallucinogenic effects and temporary dementia, and supposedly had very strict toxicity and lethality limits (even though targeted people who became incapacitated too quickly could have become fatalities as a result of hypothermia caused by belladonna-type poisoning). This plan was authorized by Project Officer Wouter Basson and initially placed under the direction of the head of Delta G's CBW division, Dr. Gert Lourens, but it was subsequently inherited by Dr. Johan Koekemoer in August or September of 1989, after he had succeeded Lourens in that position. The South African variant of BZ also included a Carboxy-Methoxy-Benzoxytropane (abbreviated as CB) in the mixture, a variant of cocaine reportedly intended to stabilize the compound and reduce the resulting aggression levels. Small amounts of this variant were then sent to RRL for animal testing, but the results were unsatisfactory because the dosage given to the test animals was either too low and caused no effect or was so high that it triggered psychosis. Basson's lawyer Jaap Cilliers further claimed that some BZ was used to block brain cells in connection with the peptide research carried out in the laboratories at Special Forces (SF) headquarters. The rest was stored, under military guard, in a storeroom at the South African Medical Services (SAMS) depot on Dequar Road along with CR, cocaine, and Ecstasy, to which only Basson himself had a key.

Alas, considerable confusion remains about the actual origins and quantities of BZ available in South Africa. Although Koekemoer, Delta G researcher Dr. Hennie Jordaan, and Delta G managing director Philip Mijburgh all later testified that they believed that only relatively small "laboratory" – i.e., gram-size – quantities of BZ were synthesized at their facility, Basson himself claimed that 1,000kg of the agent were produced at Delta G, and that experimental prototype models for its weaponization had also been created. He added that three tons of BZ was eventually used up in the process of determining the optimal BZ formula, burned up in pyrotechnical tests carried out at the Pilot Plant at SF HQ, or weaponized by being added to hand grenades, 81mm mortars and 155mm projectiles. If so, such a large quantity must have either been manufactured at facilities in South Africa or acquired overseas. Basson and his attorneys later insisted that four or five tons of BZ had been purchased in Hong Kong in April and May of 1992 from the Pharma 150 company – the "pharmaceutical" facility in the Libyan town of Rabta that is believed to have produced CW agents – with the assistance of Libyan intelligence operative `Abd al-Razaq, but prosecutors and forensic auditors considered this to be a bogus cover story concocted to explain what had happened to over 2 million dollars in unaccounted for SADF funds that may have been misappropriated by Basson. If Basson did in fact lie about this, these stocks of BZ clearly must have been obtained elsewhere or produced in-country. Whatever their source, by November 1992 all that remained after testing and weaponization was 980kg of BZ ("Product B"), 21 green drums full of which Basson claimed he dumped into the sea along with other dangerous chemicals in January 1993. The destruction of this BZ and the remaining stocks of illegal drugs that had been allegedly intended for use as "calmatives" (such as methaqualone and cocaine), which was personally "certified" by Basson but never confirmed, was required in order to bring South Africa into conformity with the terms of the Chemical Weapons Convention.

The South Africans may have even used BZ once as a battlefield weapon. On 16 January 1992 the SADF allegedly tested an unspecified chemical weapon – possibly BZ – by using a pilotless observer aircraft to bomb FRELIMO troops operating near Ngungwe on the Mozambican border, an attack that killed at least five and injured ten others. According to various sources hostile to the apartheid government, the South Africans then sought to blame this same CW attack on the African National Congress (ANC), and even went so far as to send in a team consisting of Basson, Jan Lourens, Dr. Brian Davey, and other colleagues from SAMS' 7 Medical Battalion Group to "investigate." Although Knobel and Mijburgh later denied knowing anything about such an attack, and the former insisted that teams of international investigators were unable to confirm its occurrence, at least one British scientist from Porton Down was convinced it had taken place and both the U.S. and British governments issued diplomatic démarches, suggesting that they strongly suspected South Africa's involvement. Critics of the government later accused Basson, Brigadier Van Wyck, and Colonel At Nel, justifiably or not, of being the masterminds behind the attack. They further alleged that the small unmanned aircraft used in the attack had first been tested at Komatipoort airport, and that the poison it disseminated had been manufactured and stored at Protechnik, one of the many Coast-linked companies. None of these claims have been confirmed.

In 1990 President F. W. De Klerk prohibited the carrying out of any further work on lethal CBW agents, but permitted the continued production and testing of Riot Control Agents or incapacitants such as CS, CR, and BZ. Basson himself claimed that the production of these riot control agents was actually accelerated in the period leading up to the 1993 signing of the CWC, the same year that Project Coast was officially terminated. The stocks of CW agents produced by South Africa were supposedly destroyed in conformity with international agreements, despite the fact that the actual destruction process was never independently verified. Although several laboratories there continue to produce highly toxic substances for normal industrial and agricultural use, none of these substances appear to be intended for deployment as lethal anti-personnel agents. The current government still has access to the type of technical expertise and the sort of sophisticated R&D facilities that would enable it to initiate a new CW program, but there is little reason to suppose that it has any interest in doing so.

	Updated February 2006