Agricultural statistics has played an important and significant role in the development of linear mixed models.  Today, mixed models are nearly ubiquitous in every discipline and form of statistical analysis, whether it is based on experimental design, on sampling design, or on modeling. Many advances and new developments in mixed modeling have bearing on the analysis of agricultural data. The aim of this presentation is to highlight some of these developments and to demonstrate their applicability and importance in agronomy. For example, the presentation touches on

mixed model smoothing  (e.g., spatial prediction in large data sets with smoothing splines)

the analysis of non-normal data (e.g., split-plot designs with binary data, ordinal data, or counts)

statistical graphics for treatment comparisons (e.g., all pairwise comparison displays)

the analysis of interactions (e.g., simple effects differences)