Cotton wilt of physiological origin appears to have been present in the Gezira area of the Anglo‐Egyptian Sudan since 1924 if not previously. The present investigations showed that between mid‐October and mid‐December three factors occur which cause water stress in the crop. At some time during this period there is marked reduction in the size of the absorbing system of the plant, correlated with maximum boll development. Greatly increased day temperatures and higher evaporation prevail, while during the whole period available water in the soil declines. The hypothesis is put forward that when all three factors occur at the same time at their maximum intensity, permanent wilting and death of the plants result; at lesser intensities, wilting is followed by recovery, but with reduction of yield. Experiments are described, the results of which appear to confirm this view. In the author's opinion a return to the earlier sowing date formerly favoured locally would lead to considerably increased l