The purpose of this randomised controlled trial was to examine the effect of regular exercise (three times/week) performed during the second and third trimester of pregnancy on the risk of prepartum maternal anaemia (haemoglobin <11 g/dl). 160 sedentary healthy women were randomly assigned either to a training or a control group (n = 80 each). Haematological variables were recorded at baseline (first trimester) and the beginning of the second and third trimesters. The frequency of participants with maternal anaemia was similar in the two groups over gestation (and consistently <10%). The course of haematological adaptations was normal and similar in both groups, reflecting a steady fall in haemoglobin concentration over the second trimester due to haemodilution and a rise later in gestation due to haemoconcentration. In summary, regular exercise during the second half of pregnancy does not increase the risk of maternal anaemia nor does it alter haematological variables.