This study reports an analysis of 176 newspaper articles about cancer clusters. The type of diseases, number of diseases, and status of ongoing health investigations were reported in the majority of articles. But background information such as, comparison of actual and expected number of cases, environmental causes and their confounders, and the existence of other clusters were not reported as often. Statements about risk were reported in most articles and, contrary to our expectations, were relatively balanced among risk asserting, risk denying, and mixed opinions. We noted a preponderance of risk asserting statements in the beginning of stories. Cluster stories cited an average of three sources, and two of these three sources were usually government officials. The remaining sources were divided among citizens, representatives of environmental advocacy groups and industry, and unaligned experts. Drawing on these data, we offer some suggestions to improve communication between scientists and journalists.