The flux plate method is the most commonly employed method for measuring soil heat flux (G) in surface energy balance studies. Popularity of this method is attributed to readily available sensors from several commercial sources and ease in making continuous in-situ measurements. Nonetheless, significant errors in G measured with flux plates can occur unless proper calibration and installation techniques are used and necessary corrections made. The objective of this research was to quantify potential errors in measured G when using soil heat flux plates of contrasting designs (thermal properties and dimensions) and under varying environmental conditions. Five different designs of flux plates with thermal conductivity ranging from 0.23 to 1.0 W/m K, area from 4.9 to 50 sq. cm., and thickness from 2.6 to 7 mm were evaluated. Laboratory and field experiments were completed to compare plate performance under conditions of varying G and soil water content. Measured G were also compared with G determined using gradient and calorimetric techniques. Results indicate that plate performance was significantly affected by the difference between plate and soil thermal conductivity. Observed heat flow divergence errors are compared with predicted errors calculated using available theory. Strategies for addressing errors associated with heat flow divergence around plates and real time corrections for heat storage above plates are discussed.