The general theme of this dissertation is the study of impacts of weather variability on crop yields, with each chapter addressing a specific topic related to this theme. Chapter 2 tests the hypothesis that corn and soybeans have become more drought tolerant by regressing county yields on a drought index and time. Results indicate that corn yield losses from drought of a given severity, whether measured in quantity terms or as a percentage of mean yields, have decreased over time. Soybean percentage yield losses have also declined but absolute losses have remained largely constant. The potential impact of increased drought tolerance on U.S. crop insurance rates is illustrated by comparing Group Risk Plan (GRP) premium rates assuming time-invariant susceptibility to drought with rates generated from regression results in this dissertation. Chapter 3 develops a linear spline model with endogenous knots to capture the non-linear impacts of rainfall and temperature on corn yields. A hierarchical structure is applied to capture the county-specific factors determining corn yields. Using Bayesian techniques, the thresholds and other model parameters are simultaneously estimated. Gibbs sampling and the Metropolis-Hastings algorithm are applied to estimate the posterior distributions. Corn yield decreases significantly above the upper temperature threshold and below the lower rainfall threshold. Results indicate a geographically clustering pattern of how corn yields respond to changes in temperature and rainfall. Chapter 4 applies the linear spline yield model developed in chapter 3 to examine weather impacts on yield trend, yield risk, and the distribution of corn yield. The climate trend from 1980 to 2009 explains up to 20% of observed yield trend. Not controlling for temporal weather patters leads to biased trend estimates, especially for short times series. Isolating changes in weather variability in the sample period, the hypothesis of constant coefficient of variation is rejected in most states in the Corn Belt. Decreasing marginal benefit of weather partly explains why corn yield is negatively skewed. Conditional on weather, the distribution of unexplained residuals from our yield model is symmetric in general.