WHO recommends that all pregnant women be screened for anaemia. In rural Africa this is often done by clinical examination which is known to have variable reliability. The recently developed WHO Haemoglobin Colour Scale may be the answer to this problem as it is simple and reliable. This study examines the training procedure recommended by WHO for the Haemoglobin Colour Scale when resources are very limited. We trained 7 laboratory technicians from the Medical Research Council Laboratories Hospital, Fajara, The Gambia and 13 Community Health Nurses (CHNs) from North Bank Division East, a rural area in The Gambia, to use the Colour Scale. The CHNs used the Scale to estimate haemoglobins on all new bookings to the antenatal clinics for a period of one month and recorded how they were managed. At the end of the study period they completed a qualitative questionnaire about the scale. Both groups of trainees were successfully trained although the WHO protocol for training was impossible to follow due to resource limitations. Eight of the 13 trained CHNs used the scale in practice and recorded 307 estimations with a mean haemoglobin of 9.1 g/dl. The results were normally distributed. Six of the 9 patients with Hb readings of < 4 g/dl were managed correctly. In response to the questionnaire the CHNs thought the scale was cheap, easy and quick to use and as good as the haemoglobinometer they had used previously. The main criticism was that it was not robust enough. The development of a low-technology, cheap, simple and reliable method for measuring haemoglobin is a welcome development. However, a simpler training procedure and a standard way of measuring observer performance are necessary.