The dangers of dipstick research are well known. Ask a group of men standing around a TV set in a pub in Johannesburg what they think of gewurztraminer and you'll get a response that would be dramatically different from a comparable crowd in a bar in Alsace.

If you approach what you think is the 'average' South African white wine drinker for his/her impression of what are the most popular white grape varieties, what you will hear will probably not reflect the reality of current grape prices. The smart end of the market believes that the demand for sauvignon blanc continues unabated. That's what most retailers will tell you. (Incidentally, based on this impression, sauvignon blanc plantings continue to grow).

In the six years from 2003 to 2009 the bulk price for sauvignon grapes increased from an index of 100 to 140, a pattern which does little more than track inflation. Still, that's a whole lot better than merlot (down from 100 to 68) and cabernet sauvignon (hovering at 72). However, it's no match for chenin blanc, whose seemingly inexorable price rise now sees its index at about 160.

This is not a figure that could have been predicted from chatting to retailers. Chenin has been growing off quite a low base and much of what is being bought goes into nondescript blends - so how could they be expected to know what kind of pressure on fruit prices occurs in this segment?

They might be equally surprised by how the Chardonnay price has held up. It's become so fashionable to repeat the "anything but chardonnay" mantra that no one really looks at what has been happening in this market. In many ways, chardonnay should be tracking red wine: it is generally vinified in the same way and its consumers are more tolerant of the evidence of oak maturation. However, unlike the fashionable red cultivars - whose price indices are all down - chardonnay is 20% up on the 2003 trend and still growing.

I suspect that some of this robust resistance to downward pricing pressure lies in the growers themselves being victims of dipstick research. Confident that the demand has been tapering off, they've been under-planting for years. The index might be holding fast because of a genuine shortage of fruit.

It is certainly clear - from the wines I've tasted recently as well as from South Africa's international show performance - that the variety is generally being managed by people who care about what goes into the bottle. I'm not suggesting that there isn't the same cynical exploitation that certainly goes on with sauvignon blanc - just that there's less of it.

La Motte's 2008 Chardonnay is a really good wine, with citrus and tropical notes, rich leesy creaminess and well-integrated oak. Eikendal's 2008 is equally impressive: more limey, generally fresher and less opulent, it's a very linear style and may even appeal to sauvignon drinkers. Of course, if they want to avoid the impression of oak entirely, the 2008 Middelvlei (unoaked) Chardonnay is a fabulous example.

Still, the best of my recent chardonnay notes is the 2007 Dombeya. It is now showing lovely development, with tropical apricot notes, a hint of grilled hazelnuts and some quite funky aromatics. At three years of age, it's not old, but it is comfortably poised on the plateau of maturity.

I then looked at a set of chenin blanc tasting notes - and discovered that only one of the wines in the comparably long line-up had shown any real merit. This was the Bellingham "The Bernard" Old Vines Chenin 2008. Here - like the best of the chardonnays - there was real fruit intensity and ample vinosity. It's a big wine, and perhaps a bit imposing (a euphemism for a little too much alcohol). But it also helps to explain why Chenin Blanc is out-performing the index. There aren't many bottles which offer all this flavour and complexity at that price point.

From Business Day, 3 March 2010