Paddlefish Polyodon spathula (n = 576) were collected from Kentucky Lake, Kentucky-Tennessee, with experimental gill nets in 2003-2004 to assess population characteristics and the potential for commercial overfishing. Additional data were collected from 1,039 paddlefish caught by commercial gillnetters in this impoundment. Since the most recent study in 1991, size and age structure have been reduced and annual mortality has tripled. In the 1991 study, 37% of the fish collected were older than the maximum age we observed (age 11), and in 2003 annual mortality for paddlefish age 7 and older was high (A = 68%). Natural mortality is presumably low (<10%) for paddlefish; therefore, exploitation in recent years is high. Estimates of total annual mortality were negatively related to river discharge in the years preceding each estimate. The number of paddlefish harvested since 1999 was also negatively related to river discharge because gill nets cannot be easily deployed when discharge exceeds approximately 850 m3/s. Large females spawn annually because all females longer than 1,034 mm eye-fork length (EFL) were gravid. No mature females were protected by the current 864-mm minimum EFL limit. At a low natural mortality rate, higher size limits when exploitation was high (40-70%) increased simulated flesh yields by 10-20%. Even at low levels of exploitation (21%), spawning potential ratios (SPRs) under the current 864-mm minimum EFL size limit fell below 20%. If the size limit was raised to 1,016 mm EFL, the population could withstand up to 62% exploitation before the SPR falls below 20%. An analysis of annual mortality caps indicated that the best way to increase the average size of harvested fish is to increase the minimum size limit. Recruitment overfishing probably occurs during drought years; however, variation in river discharge has prevented the population from being exploited at unsustainable rates in the past. ?? Copyright by the American Fisheries Society 2005.