This work aims at applying the theory of the component-based normalization for the Jagiellonian PET scanner, currently under development at the Jagiellonian University. In any Positron Emission Tomography acquisition, efficiency along a line-of-response can vary due to several physical and geometrical effects, leading to severe artifacts in the reconstructed image. To mitigate these effects, a normalization coefficient is applied to each line-of-response, defined as the product of several components. Specificity of the Jagiellonian PET scanner geometry is taken into account. Results obtained from GATE simulations are compared with preliminary results obtained from experimental data.