Measurements of return stroke currents on instrumented towers reveal the existence of reflections at the tower bottom and top. In the paper, the effects of the reflections on both measured currents and the original lightning electromagnetic fields are discussed. To this purpose, an extension of the modified transmission line (MTL) return stroke model is given. The model takes into account multiple reflections at the bottom and at the top of the object and allows for the `decontamination' of measured lightning currents from reflections due to the presence of the strike object. The authors' results show that even though the reflections might not be discernible in experimental recordings for tower heights as those used in Berger et al.'s (1975) and Garbagnati and Lo Piparo's (1982) investigations, the peak current might be significantly affected. Therefore, consideration of the reflections in the elaboration of meaningful statistics on lightning current is required, as well as an accurate evaluation of the tower reflection coefficients. Radiated fields calculated considering and neglecting the presence of the strike object differ. However it is shown that relatively short strike objects, such as triggered lightning launching structures, do not appreciably affect the calculated field peak values and overall wave shapes, but only the field derivatives