Volatilization of lindane from bare soil and from a closed sugar beet canopy was investigated in nine field trials. Principally two independent methods of volatilization estimation were at our disposal: the explicit calculation from vertical concentration gradients in the air based on micrometeorological theory of turbulent mass flux (we call it the direct method), and the indirect method of measuring residue disappearance, However, the indirect method could not be applied in the bare soil experiments owing to large scatter in the residue data, The use of small sized target areas required a correction of the common flux-gradient relationships (direct method). Using a well-known analytical solution of the two-dimensional mass balance equation we developed a new correction method, The cumulative volatilization results obtained over sugar beet by the corrected direct method agreed satisfactorily with the results of the indirect method. Thus we could assume the corrected direct method to work well also over bare soil. At present the direct method seems to be the only approach which yields reliable estimates of lindane volatilization from bare soil. We found pesticide concentrations and volatilization rates to be remarkably influenced by micrometeorological conditions. This effect is the more intense the higher volatilization potentially is. In general, over sugar beet the cumulative relative volatilization tended to 80% and more after a few hours while over bare soil it was observed to be less than 30% even after 2 days.