This paper investigated the porosity controlling factors for tight sandstone reservoir in the Daniudi gas field, Ordos Basin based on an integrated petrographic, petrophysical and geostatistical analyses, and proposed a comprehensive prediction model for reservoir porosity. Compaction was found to be a key factor for causing reservoir densification. The degree of sandstone compaction appears to be affected by grain sizes and sorting. Under normal compaction conditions(e.g., cement content less than 6%, and with no dissolution), the variation in reservoir porosity with burial depth can be well correlated with grain compositions, grain sizes, and sorting. Based on qualitative examination of the controlling factors for reservoir porosities, geostatistics were used to quantify the effects of various geological parameters on reservoir porosities. A statistical model for comprehensive prediction of porosity was then established, on the assumption that the present reservoir porosity directly relates to both normal compaction and diagenesis. This model is easy to use, and has been validated with measured porosity data. The porosity controlling factors and the comprehensive porosity prediction can be used to quantify effects of the main controlling factors and their interaction on reservoir property evolution, and may provide a reference model for log interpretation. This paper investigated the porosity controlling factors for tight sandstone reservoir in the Daniudi gas field, Ordos Basin based on an integrated petrographic, petrophysical and geostatistical analyzes, and proposed a comprehensive prediction model for reservoir porosity. Compaction was found to be a key factor for causing The degree of sandstone compaction appears to be affected by grain sizes and sorting. Under normal compaction conditions (eg, cement content less than 6%, and with no dissolution), the variation in reservoir porosity with burial depth can be well correlated with grain compositions, grain sizes, and sorting. Based on qualitative examination of the controlling factors for reservoir porosities, geostatistics were used to quantify the effects of various geological parameters on reservoir porosities. A statistical model for comprehensive prediction of porosity was then established, on the assumption that the present reservoir porosity directly The model is easy to use, and has been validated with measured porosity data. The porosity controlling factors and the comprehensive porosity prediction can be used to quantify effects of the main controlling factors and their interaction on reservoir property evolution, and may provide a reference model for log interpretation.