Estimating with a high precision the number of agricultural holdings and other population-related indicators is one of the main targets of the agricultural surveys. Achieving this objective is not straightforward when using an area frame: the choice of the estimator is indeed delicate since it depends on the criterion to link the observation units (e.g. agricultural holdings or tracts) and the sampling units (e.g. segments or points). According to the rule chosen, different types of estimators are defined: the open segment estimator, the closed segment estimator, the weighted segment estimator, the Horvitz–Thompson estimator and the multiplicity estimator. Moreover, auxiliary variables correlated to the population indicators can be used to produce the estimates: this is the case of the ratio estimator. In this paper, describe the estimators mentioned above, focusing particularly on the comparison between them when the target parameter to be estimated is the number of agricultural holdings.