When children are admitted to hospital, they have to adapt to new and unfamiliar stimuli. Children may respond with fear to stimuli such as pain or unfamiliar experiences. One goal of nursing is to help hospitalized children to adapt effectively to their hospital experience. Accordingly, nurses need to assess childrens' fears of their hospital experience to contribute to the planning of care to alleviate these fears. The problem addressed by this study was to identify and measure hospital-related fears (hereafter called HRF) in hospitalized school-aged children. The study was conceptualized with Roy's model. A descriptive qualitative approach was used first, followed by a quantitative approach. This study was conducted from November 30, 1989 to January 12, 1991. The sample consisted of 395 hospitalized school-aged children selected through an allocated sampling technique in nine general hospitals. The HRF questionnaire (three point likert scale) was developed by a delphi technique. The data were analyzed by an SAS program. Factor analysis was used for the examination of component factors. Differences in the HRF related to demographic variables were examined by t-test, analysis of variance and the Scheffe test. The crude scores of the HRF scale were transformed into T-scores to calculate the standard scores. The results included the following: 1. Forty-four items were derived from 188 statements identifying the childrens' hospital-related fears. These items clustered into 14 factors, fear of injections, operations, bodily harm, others' pain, medical rounds, physical examinations, medical staff, disease process, blood and X-rays, drugs and cochroaches, tests, harsh discipline from parents or staff, being absent from school, and separation from family. The 14 factors was classified into four categories, 'pain', 'the unfamiliar', 'the unknown' and 'separation'. 2. The reliability of the HRF instruments was .92 (Cronbach's alpha). In the factor analysis, Cronbach's alpha coefficients for the 14 factors ranged from .84 to .86 and Cronbach's alpha coefficients for the four categories ranged from .70 to .84. Pearson correlation coefficient scores for relationships among the 14 factors ranged from .11 to .50, and among the four categories, from .44 to .63, indicating their relative independence. 3. The total group HRF score ranged from 45 to 130 in a possible range of 44 to 132, with a mean of 74.51. The fears identified by the children were, in order, injections, harsh discipline by parents or staff, bodily harm, operations, medical staff, disease process, and medical rounds; the least feared was others' pain. The fear item with the highest mean score was surgery and the lowest was examination by a doctor. HRF scores were higher for girls than for boys, and for grade 1 students than for grade 6 students. HRF scores were lower for children whose fathers were over 40 than for those whose fathers were in the 30 to 39 age group, and whose mothers were over 35 than for those whose mothers were in the 20 to 34 age group. HRF scores were lower when the mother rather than any other person stayed with the child. The expressed fear of pain, the unfamiliar, the unknown and of separation directs nurses' concern to the threat felt by hospitalized children to their concept of self. This study contributes to the assessment of fears of hospitalized children and of stimuli impinging on those fears. Accordingly, nursing practice will be directed to the alleviation of pain, pre-admission orientation to the hospital setting and routines, initiation of information about procedures and experiences and arrangments for mothers to stay with their children. Recommendations were made for further research in different settings and for development and testing of the instrument.