A new kit for screening for sickle cell haemoglobin (haemoglobin S) has been evaluated. The kit is based on the principle that cells containing haemoglobin S, which have been induced to sickle by exposure to a reducing agent in a phosphate buffer, do not pass through a gel contained in microtubules. In theory, they can thus be distinguished from cells that do not contain haemoglobin S. In practice, the kit was found to be insensitive and unreliable for the detection of haemoglobin S. Homozygosity and heterozygosity for haemoglobin S could not be reliably distinguished. Similarly, patients with haemoglobin S who had been transfused could not be distinguished reliably from those who had not been transfused. In conclusion, unless improved performance can be demonstrated, use of this kit cannot be recommended.