Key to radiation therapy is to apply a high tumor-destroying dose while protecting healthy tissue, especially near organs at risk. To optimize treatment for ion therapy not the dose but the dose multiplied by the relative biological effectiveness (RBE) is decisive. Proton therapy has been based on the use of a generic RBE, which is applied to all treatments independent of dose/fraction, position in the spread-out Bragg peak (SOBP), initial beam energy or the particular tissue. Dependencies of the RBE on various physical and biological properties are disregarded. The variability of RBE in clinical situations is believed to be within 10-20%. This is in the same range of effects that receive high attention these days, i.e., patient set-up uncertainties, organ motion effects, and dose calculation accuracy all affecting proton as well as conventional radiation therapy. Elevated RBE values can be expected near the edges of the target, thus probably near critical structures. This is because the edges show lower doses and, depending on the treatment plan, may be identical with the beam's distal edge, where dose is deposited in part by high-LET protons. We assess the rationale for the continued use of a generic RBE and whether the magnitude of RBE variation with treatment parameters is small relative to our abilities to determine RBE's. Two aspects have to be considered. Firstly, the available information from experimental studies and secondly, our ability to calculate RBE values for a given treatment plan based on parameters extracted from such experiments. We analyzed published RBE values for in vitro and in vivo endpoints. The values for cell survival in vitro indicate a substantial spread between the diverse cell lines. The average value at mid SOBP over all dose levels is approximately 1.2 in vitro and approximately 1.1 in vivo. Both in vitro and in vivo data indicate a statistically significant increase in RBE for lower doses per fraction, which is much smaller for in vivo systems. The experimental in vivo data indicate that continued employment of a generic RBE value of 1.1 is reasonable. At present, there seems to be too much uncertainty in the RBE value for any human tissue to propose RBE values specific for tissue, dose/fraction, etc. There is a clear need for prospective assessments of normal tissue reactions in proton irradiated patients and determinations of RBE values for several late responding tissues in animal systems, especially as a function of dose in the range of 1-4 Gy. However, there is a measurable increase in RBE over the terminal few mm of the SOBP, which results in an extension of the bio-effective range of the beam of a few mm. This needs to be considered in treatment planning, particularly for single field plans or for an end of range in or close to a critical structure. To assess our ability to calculate RBE values we studied two approaches, which are both based on the track structure theory of radiation action. RBE calculations are difficult since both the physical input parameters, i.e., LET distributions, and, even more so, the biological input parameters, i.e., local cellular response, have to be known with high accuracy. Track structure theory provides a basis for predicting dose-response curves for particle irradiation. However, designed for heavy ion applications the models show weaknesses in the prediction of proton radiation effects. We conclude that, at present, RBE modeling in treatment planning involves significant uncertainties. To incorporate RBE variations in treatment planning there has to be a reliable biological model to calculate RBE values based on the physical characteristics of the radiation field and based on well-known biological input parameters. In order to do detailed model calculations more experimental data, in particular for in vivo endpoints, are needed