The daytime work load, as estimated by a nurse/patient dependency scheme, is contrasted with the hours of nursing made available in thirty different wards in three hospitals over a 3-month period. When these two sets of figures are compared it is clear that considerable disparities exist between what might be labelled as the supply of and demand for nursing care. In order to identify the source of the variability in the ratio of these two measures the data were subjected to an analysis of variance. This revealed that in two of die three hospitals most of the explained variability arose from difficulties experienced in meeting week by week fluctuations whereas in the third hospital most was attributable to difference between days of the week.