ABSTRACT We present 56Ni mass estimates for 110 normal Type II supernovae (SNe II), computed here from their luminosity in the radioactive tail. This sample consists of SNe from the literature, with at least three photometric measurements in a single optical band within 95–320 d since explosion. To convert apparent magnitudes to bolometric ones, we compute bolometric corrections (BCs) using 15 SNe in our sample having optical and near-IR photometry, along with three sets of SN II atmosphere models to account for the unobserved flux. We find that the I- and i-bands are best suited to estimate luminosities through the BC technique. The 56Ni mass distribution of our SN sample has a minimum and maximum of 0.005 and 0.177 M⊙, respectively, and a selection-bias-corrected average of 0.037 ± 0.005 M⊙. Using the latter value together with iron isotope ratios of two sets of core-collapse (CC) nucleosynthesis models, we calculate a mean iron yield of 0.040 ± 0.005 M⊙ for normal SNe II. Combining this result with recent mean 56Ni mass measurements for other CC SN subtypes, we estimate a mean iron yield <0.068 M⊙ for CC SNe, where the contribution of normal SNe II is >36 per cent. We also find that the empirical relation between 56Ni mass and steepness parameter (S) is poorly suited to measure the 56Ni mass of normal SNe II. Instead, we present a correlation between 56Ni mass, S, and absolute magnitude at 50 d since explosion. The latter allows to measure 56Ni masses of normal SNe II with a precision around 30 per cent.