Debate over civil justice reform in the United States frequently centers on the extent to which damage awards granted by juries have been escalating over time. However, past studies on civil juries have been hampered by lack of data on verdicts spanning a sufficiently long time period. Average jury awards tend to be highly variable from year to year, making it difficult to distinguish trends over relatively short periods of time. We use the longest time series of data on jury verdicts ever assembled: 40 years of data on tort cases in San Francisco County, CA and Cook County, IL collected by the RAND Institute for Civil Justice. We find that while there has been a substantial increase in the average award amount in real dollars, much of this trend is explained by changes in the mix of cases, particularly a decreasing fraction of automobile cases and an increase in medical malpractice. Claimed economic losses, in particular claimed medical losses, also explain a great deal of the increase. Although there appears to be some unexplained growth in awards for certain types of cases, this growth is cancelled out on average by declines in awards in other types of cases.