Caster Semenya: athletics facing a dilemma that strikes at the heart of gender-segregated sport

In Features & Analysis, Opinion by Jack Congdon and Richard Ollington4 Comments

It would be difficult for anyone with even a passing interest in sport to have missed the recent debate surrounding Caster Semenya, as the 800 metre runner stormed to Olympic gold in Rio this month.

Although all three of the Rio Women’s 800m medalists have been publicly questioned about having high testosterone levels, much of the media’s attention has been focused on the South African middle-distance sensation who, ever since winning the World Championships in Berlin in 2009 at the age of 18 with a remarkable time of 1:55.45, has been subjected to an unpleasant degree of highly personal interrogation and newspaper conjecture.

Image result for semenya berlin win

At the time, the IAAF had only vague policies on gender segregation, something which was changed in the aftermath of her 2009 victory. The new policy said that women could compete only if their testosterone levels were below an upper limit. That upper limit, 10 nmol/L, was set up based on a study done on all the women competing in the World Championships in 2011 and 2013. Those who exceeded this limit were made to take medication reducing their testosterone levels, and, perhaps predictably, Semenya’s times after 2009 dropped (though this also coincided with a period of injury).

The situation changed last year when the Court of Arbitration for Sport ruled that the IAAF should temporarily suspend the enforced upper limit of testosterone for two years. However, in an accompanying document running to 161 pages, they also acknowledged that “testosterone is the best indicator of performance differences between male and female athletes”, and emphasised that the ruling was made partly so that further research could determine a more scientific limit to apply. Nevertheless, Semenya, along with a handful of others, was thus cleared to compete at her previously high testosterone levels.



A rough graphical representation of hormone level distribution. Source: Joanna Harper

Prominent transgender athlete Joanna Harper has written and researched extensively on the matter of performance advantage from natural testosterone levels. Findings show that testosterone levels usually lie between 10-35 nmol/L for men and 0.4-2 nmol/L for women. Of the intersex athletes analysed Harper found their average testosterone level to be 21 nmol/L, a level which she predicts would have granted them a 10s time advantage over the course of an 800m race. Such an advantage is a game changer in an event where the difference between first and last can be just a couple of seconds.

Whether or not this advantage is fair is an argument which touches upon a number of extremely sensitive issues. It concerns a notion of womanhood seen by some to be defined rigidly by the expectations of European administrators and competitors; it involves the rights of inter-sex or trans-gender people to participate in sport without requiring medical treatment to affect their hormone levels; and it calls into question the nature and purpose of gender-segregated sport as it is presently understood.

What has made the whole debacle more difficult is misrepresentation and a tone and language of debate that has been singularly unhelpful.

There have been publications casting spurious aspersions about Semenya’s biological make-up. There have been misjudged comments by high-profile ex-athletes like Paula Radcliffe, who described the 800m event as ‘no longer sport’ when Semenya was competing.

And, most provocatively and most damagingly, there were the remarks by Poland’s Joanna Jozwik, who finished fifth between Canada’s Melissa Bishop and Britain’s Lynsey Sharp in the final, that “I’m glad I’m the first European, the second white” to cross the line.

This has given the entire issue a simplistic and dichotomous character that belies the great complexity of its content; it has allowed writers like Katrina Karkazis to cast the question about testosterone levels as one which ‘disproportionately harms black and brown women from the global south’, despite only anecdotal foundations to this claim (see the notable case of German athlete Dora Ratjen in 1936). 

To make matters worse, commentators have muddied the waters even further with observations only tangential to the question at hand. In a highly commented piece in the Guardian, Sisonke Msimang bemoans ‘the irony of athletes from Great Britain, which spent £275m on preparations for the Rio games, raising fundamental questions about fairness in a race against an athlete from a country that spent less than £1.9m has somehow been lost.’ There is undoubtedly a legitimate argument to be had about the justice of Olympic competition when funding levels are so imbalanced; but this is a quite separate issue to that of hormone levels of competing female athletes.

There is a final way in which this question has been submerged under a flood of irrelevant whataboutery, and that is the inclination of those in Semenya’s corner to raise the issue of other natural capacities which give advantage to sportspeople in different events. Jeré Longman is typical in this respect: he observes that ‘Kenyan and Ethiopian marathoners live and train at altitude, naturally enhancing their oxygen-carrying capacity. And they tend to have long, thin legs that make running more energy efficient. Kevin Durant and Brittney Griner are great basketball players in part because they are nearly 7 feet tall. Eero Mantyranta, a Finnish cross-country skier who won seven Olympic medals in the 1960s, including three golds, was found to have a genetic mutation that increased his haemoglobin level to about 50 percent higher than the average man’s.’

But such an argument is patently flawed: we do not have separate categories of basketball for both short and tall people, in the same way that cross-country skiing has not been segregated between those with varying degrees of oxygen-carrying-capacity.

Athletics, on the other hand, has been specifically divided on the question of sex in order to ensure fairness of competition. You might question the wisdom of such a perspective, and see it as a residual consequence of a world-view stuck in the early twentieth-century, but if that fairness is to be maintained, then the issue of Caster Semenya, and others with elevated testosterone levels, must be extensively discussed, however unfair it is on an athlete who has done nothing other than compete to the best of her extraordinary ability.

If it allows female athletes to compete without an upper testosterone level limit, athletics faces the real risk of closing the ranks of female competition to the tiny minority of those with elevated levels. This would be a tremendous setback in the trajectory of female sport.

A testosterone limit is clearly not perfect, and it would be a hard-hearted person indeed who could not sympathise wholeheartedly with the position of Caster Semenya and other intersex athletes like her. But a solution to this problem will never be able to fully satisfy all parties involved. Ultimately, this is not a question of finding hard and fast definitions for gender that can be universally applied. It is about coming up with a practical solution to an issue which has a profound impact within the world of athletics.

A testosterone limit – imperfect, clumsy and flawed in a number of ways – remains, for the moment, the best way of ensuring that all women have a chance of competitive success in athletics. 


  1. Donald Legget

    ROWING is one of the few sports to offer 2 categories – one for Heavyweights or Open category and the other for Lightweights where Men’s crews are limited to 70 kgs and Women’s to 57.5 kgs.
    Lightweight Rowers have to weigh in – like boxers – 2 hours before their race. They tend to sweat off before weighing in and then drink and eat after this. It is all very well if they are competing at a 2-day regatta but if their event is spread over several days at the Olympics or World Championships the athletes have to prepare themselves extremely carefully.
    This weight limit leads to some extremely close racing and was introduced in the early 1970s.

  2. Anonymous

    A testosterone limit does not allow all women to compete fairly, it is unfair on Semenya etc. I believe that is not right as she has done nothing wrong! Leave the poor woman alone and let her run. I think the only argument for forcing her to adapt her natural biology would be if you think she is not a woman, and her genitalia shows she clearly is

  3. To Anonymous

    There are so many things wrong with the above comment. Including a complete lack of argument or justification of opinions.

    – The article didn’t state that Semenya had done anything wrong, only that she was unfortunate to have been born an intersex athlete (as defined by the IAAF, of course).

    – Genitalia are not a definite indicator of gender. I thought that was accepted by the majority of people these days. You can be transgender (where the biological indicators of sex don’t match the binary gender that an individual identifies wth) or intersex (where one has a number of characteristics that don’t match with gender). No one disputes that Semenya is a female by gender – simply that she has intersex traits that confer an unfair advantage over other female athletes.

    – The existence of a separate category for female athletes is in place in order to level the playing field for those without the physiological advantages enjoyed by cis-males; the most important factor behind these advantages is an almost 100-fold increase in testosterone. I hope we can all agree on that, because the scientific body of evidence seems pretty large. Intersex athletes experience some of these very advantages that the female category exists to remove, so of course it’s important to talk about ways to make the competition fairer (however awful the media spotlight on Semenya must be).

    – Right now, there are no hard-and-fast answers, which is why we need to discuss it. Allowing competition to continue as it is would be very destructive for the sport because simply no amount of training can overcome the physiological advantage of a tenfold increase in testosterone. Making intersex athletes compete in the open category (essentially the men’s category) would put them at a disadvantage due to lower testosterone levels. Making a third category for mid-range testosterone athletes (e.g. intersex athletes and transgender athletes undergoing hormone treatments) would probably be fairest, but would have a disproportionately small pool of competitors and so might suffer from lack of competition.

    Obviously, the topic is a very delicate matter for certain world-class athletes, so we need to be very careful with our language and the nature of the discussion. But we certainly can’t avoid having it. Personally, I’m very pleased that the authors were pushing in this direction.

Leave a Comment