1. Between September 1941 and January 1942 haemoglobin estimations were made by the Haldane method on 544 women and 530 children, aDd the values obtained compared with those for similar groups investigated before the war. 2. The women investigated were grouped as follows: (a) 353 nurses and students at a London teaching hospital and its base in the country, (b) 159 workers at a factory in a small town near London, and (c) 32 housewives, mothers of children attending a hospital in the East End of London. 3. The mean haemoglobin value of the 544 women was 89·5 % and the range was from 56 to 120%. The mean is very similar to that obtained by Davidson et al. in 1937 for a group of nulliparous women from the poorest class in Aberdeen. In contrast Price-Jones's (1931) mean figure of 98·3% for 100 supposedly healthy nurses and women students, and Jenkins & Don's (1933) mean figure of 100·5% for 116 women are significantly higher than the mean of the present series. The percentage haemoglobin was below 90 in 54 %, and below 80 in 10 % of our cases. In Price-Jones's series no haemoglobin value was below 90 %, though in Jenkins and Don's series it was below 90 in 9 % and below 80 in 1 % of the cases. 4. The mean haemoglobin value of all the nurses and students was 87·2 %. The mean for those working in London was 90.6 % and for those working in the country was 84·1 %: the difference between these means is significant. The mean values of the nurses and students working at the same hospital approximated closely, as did the mean value for the auxiliary nurses to that of the professional nurses at the same hospital. 5. The mode of life and conditions of work in the two hospitals did not differ materially. A comparison of the diets of the nurses at the two hospitals showed that in London, over a period of 3 months, the average daily intake yielded 1865 calories and 63·5 g. of protein, of which 41 g. were first-class protein; whereas in the country, for the same period, the diet yielded 1662 calories daily and 46·7 g. of protein, of which 28 g. were first-class protein. It was estimated that the London group had a daily iron intake of 10.8 mg. and the country group one of 9·2 mg. 6. The mean haemoglobin values of the factory workers and the London housewives were 94·0 and 89·6% respectively. 7. A small number of nurses and students with low haemoglobin values had regular iron medication and the majority showed a satisfactory response to treatment. 8. There were in the present, series 364 children aged 6 months to 5 years. They were examined at welfare centres, diphtheria immunization clinics and day-and residentialnurseries. The curve of their mean haemoglobin values at successive ages approximates to corresponding curves of 10 years ago for British children not receiving medicinal iron. In our present series the average haemoglobin level at 6–12 months was 75·4%. Between 1 and 2 years of age the level had dropped to 72·8%, and thereafter it gradually rose, reaching 81·8% between 4 and 5 years of age. The lowest levels were found in children under 2 years of age. Since iron treatment of such infants will not only raise their mean haemoglobin level, but also halve their morbidity rate (Mackay & Goodfellow, 1931), the haemoglobin level of these untreated babies must be considered pathologically low. The incidence of severe anaemia was lower in the welfare group than in the other groups and was highest among the children in the residential nurseries. 9. The mean haemoglobin level of 90 school children aged 5–13 years was 80·3% (about 10% lower than Davidson's (1935) mean for poor children in Aberdeen), of 38 school children aged 13–15 years 89·3% and of 38 adolescent girls working at the factory already mentioned 98·7 %. 10. In all the groups of women and older children, with the exception of the small group of girls working in a factory, the mean haemoglobin values were significantly lower than those of previous groups with which we could compare them. We believe that this general lowering of the mean haemoglobin values of the older subjects and the low curves obtained for the younger children are the expression of a nutritional anaemia. We also believe that iron deficiency is one important aetiological factor, but that other factors have also played a part in the production of this anaemia. It is probable that among women and school children other dietetic deficiencies are implicated, and that in nurseries a high incidence of infection has increased the incidence and severity of an anaemia primarily due to an iron deficiency. 11. We believe that the bearing of this anaemia on the health of the nation is of sufficient importance to call both for official action at the present time, and for further investigation.