Active SAPS investigation. Know something? Call 082 465 7249 or Crime Stop 08600 10111.

Patterns, and what they can't tell you

Everything below is computed from the discovery facts on the timeline. When the facts change, these numbers change. Every reading carries its own caveat, because the fastest way to get this wrong is to mistake "where a body was found" for "what happened".

7
discoveries
10 d
average gap between discoveries
1 d
most recent gap
5/7
within 2.5 km of the R21
41 km
widest distance between two sites (#7 ↔ #6)
32, 38
known victim ages

Discoveries are bunching up

The gap between the first two discoveries was 40 days. The last two gaps were 2 and 1 days. Over 62 days, 7 bodies have been found, an average of one every 10 days, but almost all of them in the last week.

Caveat: A 'found' date is not a death date. The 10 September body was badly decomposed, so that woman likely died weeks earlier. Part of the bunching is simply that searches intensified once police went public on 12 September.

Two geographic groups, not one

4 of the 7 discoveries sit within about 3 km of Rhodesfield, next to OR Tambo airport. 5 of 7 are within 2.5 km of the R21. The two that break the pattern are Olifantsfontein (still on the R21, ~20 km north) and KwaThema (a different road network, ~25 km south-east).

Caveat: Pins are approximate, most reports say only 'near the R21' or 'in Rhodesfield'. Police have not said the KwaThema death is connected at all.

Who the victims are

All are women. Known ages: 37, 32, 38 (4 unknown). Reporting describes them as Black women in their thirties. 2 publicly confirmed as identified, 1 confirmed unidentified, 4 unclear. Only one, Elizabeth Moselakgomo, has a publicly known last-seen point: she vanished during an evening run in Rhodesfield.

Caveat: SAPS releases very little victim detail, for good reasons. Unidentified victims are the biggest gap in the public record, and one where the public can legitimately help (see Contribute).

How they were found

Every reported body was outdoors, on a roadside or in open veld, naked or partially clothed, with visible bruising. One (July) was restrained with cable ties. Police described the Olifantsfontein woman as possibly 'dumped', which implies she was killed elsewhere and moved.

Caveat: No cause of death has been made public for any victim. Everything above is what a passer-by or the police spokesperson could see, not what a post-mortem found.

Against the South African baseline

The best public research on South African series (Horning, Salfati & Labuschagne, 2015: 302 incidents, 33 offenders) found 78% of scenes outdoors and 94% of bodies left where the victim was killed; nearly half of victims were lured, often with a job offer. The Kempton Park pattern matches the outdoor/undressed baseline strongly, but 'dumped' bodies and an abduction of a runner in the open would be departures from the typical con-approach series.

Caveat: The baseline is 1953–2007 and describes SOLVED series. It tells you what is usual, not what happened here.

The July arrest changes the arithmetic

An arrest was made for the first death two days after the body was found, and reporting describes the arrested person as the victim's partner. If that case stands on its own, the 'series' starts on 24 August, not 15 July, and the count is 6, not 7.

Caveat: This site does not name or discuss suspects. An arrest is not a conviction, and a domestic-looking case can still turn out to be connected.

The numbers behind the readings

FromToDays between discoveries
#1#240
#2#714
#7#33
#3#42
#4#52
#5#61

SiteDistance from the R21 line (km)Location certainty
#1 Rhodesfield0.0± 2.0 km, Reported only as 'open veld near the R21' (Daily Maverick adds 'in Rhodesfield'). Pin is a corridor-level estimate, not an exact spot.
#2 Kempton Park (R21 corridor)0.0± 3.0 km, Reported only as 'near the R21 highway'. Pin is a corridor-level estimate.
#7 Clayville3.8± 3.0 km, Reported only as 'a river in Clayville'. Clayville lies just west of Olifantsfontein and the R21, about 18 km north of Rhodesfield; the Kaalspruit runs through it. Pin is suburb-level.
#3 Rhodesfield0.7± 1.5 km, IOL places this in the Rhodesfield area; other outlets give no location. Pin is suburb-level.
#4 Rhodesfield1.1± 0.5 km, Street-level: York Road, Rhodesfield, next to the OR Tambo airport hotels. Outlets name different hotels (Protea vs Southern Sun), both are in the same block.
#5 Olifantsfontein0.0± 3.0 km, Roughly 20 km north of the Rhodesfield cluster along the R21 (The Witness: 'approximately 20 kilometres from Kempton Park'). Pin is approximate.
#6 KwaThema24.2± 1.5 km, TimesLIVE: 'behind KwaThema Centre'. Pin is on the centre; treat as ±1 km. About 25 km south-east of the Rhodesfield cluster and NOT on the R21.

Day of week foundCount
Wednesday1
Monday3
Thursday1
Saturday1
Tuesday1

How the crimes were carried out, what is actually public

Each element of the crimes, with its status: Confirmed by SAPS means an official statement said it;Reported by media means outlets reported it citing police sources; Not public means nobody official has said anything. Read the last group carefully, it is where rumours grow.

ElementStatusWhat is knownSources
Where the bodies were leftConfirmed by SAPSOutdoors, roadside verges and open veld, mostly near the R21 around Rhodesfield; one at Olifantsfontein, one behind KwaThema Centre, and one (announced 15 Sept, found 7 Sept) in a river in Clayville, the only body found in water.The CitizenIOLThe CitizenIOL / The StarTimesLIVETimesLIVE
State of dressConfirmed by SAPSSAPS: bodies found 'half naked'. Individual reports: naked (#1, #3), partially clothed (#2, #4 in running wear, #5).BNO NewsThe Citizen
Visible injuriesConfirmed by SAPSSAPS: 'visible bruising' on the women found around Rhodesfield and at Olifantsfontein. No detail on where or how severe.The CitizenSA People
RestraintsReported by mediaVictim #1 (July) was restrained with cable ties, according to multiple outlets citing police sources. Not reported for any other victim.BNO NewsDaily Maverick
Killed where found, or moved?Reported by mediaFor #5 (Olifantsfontein) Brig. Mathe said the woman 'may have been dumped' between Sunday night and Monday morning, i.e. possibly killed elsewhere and transported. Nothing public for the others.The CitizenThe Witness
How victims were takenReported by mediaOnly one case has a public last-seen point: #4 went for an evening run on Main Road, Rhodesfield (9 Sept) and did not return. For every other victim, where and how she was approached is not public.Daily Maverick
Cause of deathNot publicNo cause of death has been released for any victim. The only post-mortem finding made public is the sexual-assault finding (see below). Words like 'strangled' appear in nobody's official statement.,
Sexual assaultConfirmed by SAPSSAPS (Lt-Gen Mosikili, reported by IOL on 15 Sept): four of the five women found by 14 September had been sexually molested; one body was too decomposed to tell. Which four is not public. Nothing on the sixth or seventh.IOL / The Star
Weapon or methodNot publicNothing released.,
Time of attackNot publicOnly #4's disappearance is timed (evening). Discovery times (early morning for #5 and #6, Saturday evening for #4) say when a passer-by found them, not when they died.The CitizenIOL / The Star
One perpetrator or several?Not publicSAPS has not linked any of the deaths. One person is in custody for #1 (reportedly the victim's partner). On 14 Sept Brig. Mathe said 'nothing tells us that we've got a serial killer' and that the focus was on whether the three later deaths then known were connected; sources told The Citizen the cases did not all share the same method. Prof. Gopal reads the pattern as one perpetrator; Dr Pistorius says a crime-scene signature, DNA and pathology will decide. None of this is a finding.SABC NewsThe CitizenEWNBriefly.co.za
Sex, age or description of any perpetratorNot publicSAPS has released no description, identikit, CCTV image or vehicle. Any 'profile' circulating online is invented.,

Why there is no "offender profile" here. Behavioural profiling needs the things in the "Not public" rows, cause of death, post-mortem findings, scene forensics, the order the victims actually died in. The SAPS Investigative Psychology Unit has those; this site does not. A profile written from news snippets would be fiction with a confident tone, and on a public site it becomes a description that a crowd matches against real men in Rhodesfield. See rule 2.

AI-assisted digest

Generated by Claude (in a Claude Code / Cowork session, on subscription, no API) from 29 sources in data/sources.json as of 15 Sep 2026 (afternoon), on 2026-09-15. Not yet reviewed. It reads the public record; it does not profile anyone.

Seven women in the police count as of the afternoon of 15 September: four in a tight cluster around Rhodesfield/R21, one 20 km north on the same highway (Olifantsfontein), one in a river at Clayville just west of it (found 7 Sept, announced 15 Sept), one 25 km south-east at KwaThema on a different road network. Police hold one person for the July case; SAPS says it is 'still not yet there' on whether one person or a group is responsible and has asked for patience. The first post-mortem finding is public: four of the five women found by 14 Sept had been sexually assaulted. The public record supports a consistent disposal pattern (outdoors, undressed, bruised, now with sexual assault) and victim profile (women in their late 20s to 30s), but not a single-offender conclusion, and the police themselves are saying so.

What changed

Discrepancies between sources

Open questions

Against the South African baseline

Horning, Salfati & Labuschagne (2015): South African series are typically outdoor (78%), victims left where killed (94%), strangers (73%), often lured with a job offer (47% con approach). Kempton Park matches the outdoor/undressed norm; transported bodies and an outdoor abduction of a runner would be departures, but nothing here is confirmed by police.

What this is not: Not a profile of an offender, not a prediction of where one lives, not a list of suspects. A reading of published facts, reviewed by the site owner before publishing.

What this site deliberately does not do

No "geographic profile". There is a well-known technique that turns crime sites into a heat-map of where an offender probably lives. It needs encounter sites (where the victim was taken), a reasonable number of them, and a trained analyst. We have a handful of discovery sites with kilometre-wide uncertainty. Publishing a heat-map from that would be junk science, and on a public website it would point a crowd at a suburb. The SAPS Investigative Psychology Unit has the real data; this site does not.

Observations are checked by a moderator before they appear. The rules.

Community observations

Approved observations about the public record, a detail one outlet reported that others missed, a discrepancy, a pattern in the published facts. Reviewed by a moderator before appearing.

None approved yet. Be the first.