In breast cancer radiotherapy, significant discrepancies in dose delivery can contribute to underdosage of the tumor or overdosage of normal tissue, which is potentially related to a reduction of local tumor control and an increase of side effects. To study the impact of these factors in breast cancer radiotherapy, a meta analysis of the clinical data reported by Mavroidis et al. (2002) in Acta Oncol (41:471-85), showing the patient setup and breathing uncertainties characterizing three different irradiation techniques, were employed. The uncertainties in dose delivery are simulated based on fifteen breast cancer patients (5 mastectomized, 5 resected with negative node involvement (R2) and 5 resected with positive node involvement (R1)), who were treated by three different irradiation techniques, respectively. The positioning and breathing effects were taken into consideration in the determination of the real dose distributions delivered to the CTV and lung in each patient. The combined frequency distributions of the positioning and breathing distributions were obtained by convolution. For each patient the effectiveness of the dose distribution applied is calculated by the Poisson and relative seriality models and a set of parameters that describe the doseresponse relations of the target and lung. The three representative radiation techniques are compared based on radiobiological measures by using the complication-free tumor control probability, P1 and the biologically effective uniform dose, D concepts. For the Mastectomy case, the average P1 values of the planned and delivered dose distributions are 93.8% for a DCTV of 51.8Gy and 85.0% for a DCTV of 50.3Gy, respectively. The respective total control probabilities, PB values are 94.8% and 92.5%, whereas the corresponding total complication probabilities, PI values are 0.9% and 7.4%. For the R2 case, the average P1 values are 89.4% for a DCTV of 48.9Gy and 88.6% for a DCTV of 49.0Gy, respectively. The respective PB values are 89.8% and 89.9%, whereas the corresponding PI values are 0.4% and 1.2%. For the R1 case, the average P1 values are 86.1% for a DCTV of 49.2Gy and 85.5% for a DCTV of 49.1Gy, respectively. The respective PB values are 90.2% and 90.1%, whereas the corresponding PI values are 4.1% and 4.6%. The combined effects of positioning uncertainties and breathing can introduce a significant deviation between the planned and delivered dose distributions in lung in breast cancer radiotherapy. The positioning and breathing uncertainties do not affect much the dose distribution to the CTV. The simulated delivered dose distributions show larger lung complication probabilities than the treatment plans. This means that in clinical practice the true expected complications are underestimated. Radiation pneumonitis of Grade 1-2 is more frequent and any radiotherapy optimization should use this as a more clinically relevant endpoint.