The paper estimates fiscal multipliers based on models of structural vector autoregression, identified by sign and narrative restrictions. Narrative restrictions enhanced the identification procedure, having narrowed the set of models in line with the fiscal multipliers’ theoretical inference. The models that were predominantly associated with positive impulse responses of output to government expenditures shocks and negative impulse responses to government re­venues shocks were chosen. Narrative sign restrictions only slightly changed the median impulse responses however wiped off outlier models induced by the randomicity of sign restrictions identification. This fostered the more accurate intervals of impulse responses and improved the estimates. In result, point estimate of revenue multiplier is lower in absolute value (–0,38) than the point estimate of expenditure multiplier (0,42). Nevertheless, taking into account, the multiplier of oil and gas revenues is greater than non-oil and gas revenues. Economic expenditures have the greatest impact on GDP during the first year whereas the least have social expenditures. The contribution of national projects to GDP was evaluated using estimated multipliers given the near-2019 economic conditions. It turned out to be slightly positive in 2019 (0,4%), then it grows and raises GDP on 4,0% in 2024 against the scenario with the absence of national projects. Thus, the average uplift to GDP growth rates is 0,67 p.p.