Kenya is one of the poorest countries in Africa. The GNP per capita in Kenya was 250$us and only 5% of the GNP was allocated to health. One can make a direct link between the health expenditure and the general health status of the population in a poor country. Vulnerable population groups such as childbearing women, children, especially the very young, and the aged are more affected by the low level of funding in the health service. Mortality rates amongst childbearing women and the very young in Kenya are partly determined by the structure of the health service, availability of primary care and the accessibility to health care of these vulnerable groups. However, the basic standards of care such as hygiene, good nursing and medical practice, could be maintained in spite of the difficulties referred to. This study focuses on the magnitude, causes and factors affecting maternal death and child death in Nakuru Provincial General Hospital, Kenya Action research methodology was adopted in this study as it was believed that introducing some basic changes in hospital practice could improve care standards in the maternity and paediatric units and could save lives in the longer term. Instituting a more accurate system of recording hospital admission and clinical records could identify causes of deaths. The information could be used to justify further changes and for health planning. Preliminary investigation revealed that the hospital records were poorly kept and that many mistakes and omissions were made. More importantly, causes of deaths were not accurately recorded. A more reliable system was adopted following a 3-month pilot data collection. Examination of the 3-month data revealed that 436 women died during labour per 100 000 births and that 160 in every 1,000 children between ages of 0-5 died. It would appear that these deaths in Nakuru Provincial General Hospital were closely related to poor facilities, low staff morale, poor hospital management, lack of clinical supervision in maternity services. A programme of improvement was implemented in August 2001.Evaluation was carried out 18 months after the action programme. The hospital is cleaner, infection control practice has improved, supervision of junior nursing and medical staff in these units has improved and more importantly patients in these clinical areas are enjoying better care. It is difficult to establish whether a reduction in mortality in these 2 clinical specialities could be achieved as factors such as funding, the health service structure, the establishment of accessible primary care and changes in cultural norms and harmful health practices all play a role in mortality rates. ?Staff in the hospital believe that the action programme has benefited the hospital, but question whether the changes can be sustained.