The objective of the study was to derive a mathematical method for calculating the entrance surface dose (ESD) from exposure factors for all tube potentials used in clinical practice and to compare the calculated ESDs (ESD(C)) with those measured (ESD(TLD)) using thermoluminescent dosemeters (TLDs). The exposure parameters of 43 patients who underwent (a) posteroanterior (PA) and lateral (LAT) chest examination (13 patients), (b) supine abdomen (10 patients), (c) erectus abdomen (10 patients), or (d) urinary tract examination (10 patients) were recorded. Patient ESD was directly measured by TLDs and calculated from exposure factors. The differences between ESD(C) and ESD(TLD) were quite small and could be explained by the uncertainties involved in both methods, in all but the PA chest examination where the ESD(C) was about 50% larger than ESD(TLD). However, in PA chest the ESD(TLD) was close to the minimum detectable dose of TLDs, questioning the accuracy of ESD(TLD). Further investigation showed that using the high tube potential technique (130 kV) in the PA chest examination resulted in very short exposure times, in the region of 4 ms. In such short exposure times, the X-ray generator operation presented stability problems that led to loss of output linearity and consequently to false calculation of ESD. The calculation method offers a reliable and cheap alternative to the measurement of ESD by TLD, provided that the exposure times are not as short as in the PA chest examinations recorded in this study, so that the output linearity with tube current-time product (mAs) is maintained.