Conservation of energy is a fundamental principle in micrometeorology; net radiation (Rn) less soil heat flux (G) should equal the sum of latent heat (LE), sensible heat (H) and heat storage if the control volume is properly defined. A wide range of eddy covariance studies have shown that that closure, (Rn-G)/(LE+H), for low vegetation is typically between 0.7 and 0.9. A study of energy balance closure was conducted using a network of three eddy covariance towers deployed on tallgrass prairie in eastern Kansas, USA. The effect of measurement error, sampling scale discontinuities [sample area of (Rn-G) v. that of (H+LE)], and post-processing procedures (rotation, footprint filters, frequency response, etc) were evaluated. Discussions will include the degree of closure one should expect when making eddy flux measurements for crops, soils, and grasses. The best sampling techniques and data processing protocols for maximizing closure will be presented.