Canopy stomatal conductance is commonly estimated from eddy covariance measurements of the latent heat flux (LE) by inverting the Penman–Monteith equation. That method ignores eddy covariance measurements of the sensible heat flux (H) and instead calculates H implicitly as the residual of all other terms in the site energy budget. Here we show that canopy stomatal conductance is more accurately calculated from eddy covariance (EC) measurements of both H and LE using the flux–gradient equations that define conductance and underlie the Penman–Monteith equation, especially when the site energy budget fails to close due to pervasive biases in the eddy fluxes and/or the available energy. The flux–gradient formulation dispenses with unnecessary assumptions, is conceptually simpler, and is as or more accurate in all plausible scenarios. The inverted Penman–Monteith equation, on the other hand, contributes substantial biases and erroneous spatial and temporal patterns to canopy stomatal conductance, skewing its relationships with drivers such as light and vapor pressure deficit.