In This study was carried out to describe and analyze experimental studies conducted in graduates nursing degree. Of 170 experimental studies conducted during the past three decades between 1962 and August 1991, 150 studies were available, including 124 master's and 26 doctoral theses. This study examined their general characteristics and detailed research methods using percentiles. The results were as follows: 1. Most of the studies adopted a quasi-experimental design. 2. The subjects of the studies were chosen by convenience sampling except for two studies whose subjects were drawn by randomization. Studies comparing experimental and control groups were in the highest proportion and the most frequent sample sizes of each group were 21 to 30 for both experimental and control groups. 3. As to measurement, physiological measures were most frequent followed by psychsociological measures and active report questionnaires. Each study, on average, adopted two kind of measurement tools. Studies in which the data collection period was of 1~2 months were in the highest proportion. 4. All doctoral theses and 67.0% of master's theses examined specific research hypotheses. Of these studies, the results of 92.5% supported the hypotheses. 5. Parametric statistics were the major analytical methods. In particular, t-test was used most frequently followed by Chi square, F test, and Pearson Correlation Coefficients. 6. Patients were the most frequent study subjects. Frequent nursing interventions were information and education followed by support, distraction, and nursing treatments. 7. With regards to the dependent variables, "feelings" such as anxiety, pain, and depression were most frequent. In addition, "exchanging" such as restoring, metabolism, cardiopulmonary function, infection and vital signs were adopted as the dependent variables in 29.1% of the studies examined, while 12.3% of the studies selected "choosing" such as stress, health behavior, or role performance.