Malnutrition is the fourth commonest reason for hospital admission to the paediatric department of the Central Hospital, Maputo and has the second highest death rate (20 per cent). A study from 1995 into mortality at this paediatric department, suggested an increase in severe malnutrition. Recent studies have shown that the global burden of undernutrition in the world is declining; however, data for Eastern Africa shows a deterioration. The current study was aimed at describing and comparing the patients on the malnutrition ward, in 2001 and 1983. The study gathered indices of nutritional status and secondary diagnoses from the notes of all children (aged between 6 months and 5 years) discharged from the malnutrition ward for a period of l year (January-December 2001), and from data (collected in January-December 1983) for the malnutrition ward. Data was entered and analysed using Epi-Info 6 and SPSS statistics package. The ethics committee of the hospital approved the study. Data was collected for 558 children in 2001 and 833 in 1983. There was no gender difference, average age was 21.7 months in 2001 and 23.8 months in 1983 and the average hospital stay was 13.1 and 14.3 days, respectively. In 2001, 33 per cent had kwashiorkor, 26 per cent marasmus, and 28 per cent marasmic kwashiorkor. Three hundred and twenty children (82 per cent) were <2 Z-scores below the median weight-for-age and 252 children (65 per cent) were <3 Z-scores. Forty per cent had malaria, 65 per cent anaemia, 53 per cent bronchopneumonia, 14 per cent TB, 36 per cent diarrhoea, and 12 per cent HIV/AIDS. In 1983, 49 per cent had kwashiorkor, 17 per cent marasmus, and 11 per cent had marasmic kwashiorkor. A total of 494 children (81 per cent) were <2 Z-scores below the median weight-for-age and 335 children (55 per cent) were <3 Z-scores. Eighteen per cent had malaria, 37 per cent anaemia, 28 per cent bronchopneumonia, 6 per cent TB, 8 per cent diarrhoea, and 4.4 per cent measles/post-measles. A comparison between the clinical status of 1983 with that of 2001 shows little difference in age, gender or length of stay. There were fewer admissions in 2001, although a higher percentage of severely underweight children and the 2001 group had more secondary infections, especially malaria, bronchopneumonia and anaemia. Clinical malnutrition at a referral hospital level, in spite of the remarkable Mozambican economic growth, shows signs of following the depressing pattern for much of Eastern Africa. A prospective study including HIV tests and anthropometric data for this and the city's other hospitals is warranted. Discussion needs to be prompted on a local level about malnutrition and the use of guidelines.