This article measures the job-finding, separation, and job-to-job transition rates in the United States from 1948 to 2004. The job-finding and job-to-job transition rates are strongly procyclical and the separation rate is nearly acyclical, especially since 1985. The author develops a simple model in which unemployed workers search for jobs and employed workers search for better jobs. The model predicts that an increase in either the job-finding rate or the separation rate raises the job-to-job transition rate, which is qualitatively and quantitatively consistent with the available evidence. In contrast, if the job-finding rate were acyclical and the separation rate countercyclical, as is the conventional wisdom, the model predicts that the job-to-job transition rate would be counterfactually countercyclical.