Radiation therapy is often delivered by multiple sequential treatment plans. For an accurate radiobiological evaluation of the overall treatment, fractionation corrections to each dose distribution must be applied before summing the three-dimensional dose matrix of each plan since the simpler approach of performing the fractionation correction to the total dose-volume histograms, obtained by the arithmetical sum of the different plans, becomes inaccurate for more heterogeneous dose patterns. In this study, the differences between these two fractionation correction methods, named here as exact (corrected before) and approximate (after summation), respectively, are assessed for different cancer types. Prostate, breast, and head and neck (HN) tumor patients were selected to quantify the differences between two fractionation correction methods (the exact vs the approximate). For each cancer type, two different treatment plans were developed using uniform (CRT) and intensity modulated beams (IMRT), respectively. The responses of the target and normal tissue were calculated using the Poisson linear-quadratic-time model and the relative seriality model, respectively. All treatments were radiobiologically evaluated and compared using the complication-free tumor control probability (P+), the biologically effective uniform dose (D) together with common dosimetric criteria. For the prostate cancer patient, an underestimation of around 14%-15% in P+ was obtained when the fractionation correction was applied after summation compared to the exact approach due to significant biological and dosimetric variations obtained between the two fractionation correction methods in the involved lymph nodes. For the breast cancer patient, an underestimation of around 3%-4% in the maximum dose in the heart was obtained. Despite the dosimetric differences in this organ, no significant variations were obtained in treatment outcome. For the HN tumor patient, an underestimation of about 5% in treatment outcome was obtained for the CRT plan as a result of an underestimation of the planning target volume control probability by about 10%. An underestimation of about 6% in the complication probability of the right parotid was also obtained. For all the other organs at risk, dosimetric differences of up to 4% were obtained but with no significant impact in the expected clinical outcome. However, for the IMRT plan, an overestimation in P+ of 4.3% was obtained mainly due to an underestimation of the complication probability of the left and right parotids (2.9% and 5.8%, respectively). The use of the exact fractionation correction method, which is applying fractionation correction on the separate dose distributions of a multiple phase treatment before their summation was found to have a significant expected clinical impact. For regions of interest that are irradiated with very heterogeneous dose distributions and significantly different doses per fraction in the different treatment phases, the exact fractionation correction method needs to be applied since a significant underestimation of the true patient outcome can be introduced otherwise.