The article analyzes the approaches of state statistical bodies to the construction of analytical groupings of private subsidiary farms by the total land area and the total sown area. Analytical groupings published according to the national agricultural censuses, as a rule, represent distribution series, where the characteristics of the selected groups are limited only by the number of farms, the volume of the phenomenon and its ratio to the total. In addition, the program for publishing the results of the agricultural microcensus does not provide for a regional section. In order to improve the quality of the summary, analysis and presentation of the data on private subsidiary farms, it is proposed to revise the intervals for a number of grouping indicators and to significantly expand the set of features characterizing analytical groups. On the basis of depersonalized primary data of the agricultural microcensus, the article shows the possibilities of analyzing the size, specialization and intensification of agricultural production of private subsidiary farms while conducting a complete summary of the signs of the observation program. The trends of decrease in the total number of private subsidiary farms and their land and sown area revealed according to the census data indicate a reduction in the size of agricultural production of the category of households of the population, which makes it possible to raise the question of the expediency of their continuous survey and the introduction of censorship, as it is done in some foreign countries. On the basis of the data of the sample survey of private farms and the microcensus of 2021, an estimation of the value of the gross agricultural production by groups of farms according to the size of the land area was given, and it was proposed to introduce a requirement of 1000 m 2 of land area for the inclusion of private subsidiary farms in the list of respondents of agricultural censuses.