Field capacity (FC) is one of the most cited soil physical parameters and is relevant for the management of agricultural systems. In the search for alternatives to the estimation of FC, several researchers have proposed methods based on dynamic and static criteria. Evaluating these methods within a range of soil textural classes is important for prospects of their uses with higher water use efficiency. Based on the hypothesis that FC for a certain soil textural class is exclusively associated with an equivalent pore diameter, D-threshold, which separates structural and textural porosity, and that textural pores define the FC conditions, this study aimed to: 1) estimate the equivalent pore diameter (D-threshold) which defines FC in each soil textural class in situ and, with the change of density, from the alterations in its water characteristic curve - WCC; 2) estimate the water content at FC in each soil textural class in situ and, with the change of density, from the alterations in its WCC. To obtain the water content corresponding to FC in situ, instantaneous profile-type experiments were conducted in five soil textural classes. To estimate FC and D-threshold based on the change in pore volume (CPV), soil water characteristic curves were used considering two situations: before and after compaction. Correlation and regression analyses were carried out between D-threshold and sand percentage, between water content at FC and the fitting parameters α and n of the model of van Genuchten (1980), and between the value of the tangent at the inflection point of the soil WCC and the parameter α. By comparing D-threshold values obtained by the field and CPV methods, it was possible to observe that for the range of textural classes analyzed, in general, D-threshold values obtained by CPV were different from those obtained in situ. Therefore, the methodological procedure CPV does not represent the actual D-threshold of the FC found under field conditions. The values of water content corresponding to FC obtained by the CPV protocol are not due to the physical aspect conceived in the method's proposal. Thus, since these values were not due to the D-threshold, this protocol cannot be indicated because it became evident that there was no clear cause-effect relationship. It was concluded that: 1) the principle of the method of estimating FC must be consistent with processes occurring in the soil. Since FC is dependent on soil processes, and associated with a drainage rate, there is no good reason why the CPV method correctly estimates the attribute; 2) the water content at FC in a certain soil textural class is not exclusively associated with only one equivalent pore diameter, D-threshold, so the hypothesis assumed in the study has not been confirmed; and 3) protocols based on soil water dynamics should preferentially be used to estimate FC, to the detriment of methods based on static criteria.