The polarographic measurement of tissue oxygenation is one of the most widely used methods in clinical practice for the quantification of tumour hypoxia. However, due to the particular features of the electrode measuring process, the results of the measurements do not accurately reflect the tumour oxygenation. This study aimed to find a correlation between the electrode measurements and the tumour oxygenation in an attempt to improve the accuracy of the predictions regarding the response to treatment based on electrode measurements. A previously developed computer model that allows the simulation of tumour tissue and electrode measurements was used. The oxygenation of a large number of tumours with biologically relevant distributions of blood vessels was theoretically calculated. Simulations of electrode measurements allowed the comparison between the real tissue oxygenation and the results obtained with the electrode. A semi-empirical relationship between the hypoxic fraction measured by the electrode and the real hypoxic fraction in the tissue has been found. The impact of the correction of the electrode measurements in terms of predictions for tumour control probability was estimated for a few clinical examples. The range of possible true values corresponding to one measurement has also proven useful for explaining the apparently unexpected response to the treatment of some patients. The corrected hypoxic fraction which is believed to be closer to the real value of tissue hypoxia predicts much smaller control probabilities than the raw electrode measurements. This could provide an explanation for the apparently unexpected failure to respond to the treatment of some of the patients with apparently favourable tumour oxygenation. This also means that the electrode measurements cannot be used directly for the quantitative modelling of tumour response to the treatment. The conversion method proposed in this paper might however strengthen the statistical power of the correlations between the electrode measurements and the treatment outcome.