We have attempted to confirm previous reports of "egg transformation" in seven Nicotiana species, including genetic stocks obtained from the original experimenter. The methods employed in the original experiments were duplicated as closely as possible. In total, 1,622 pollinations were made involving irradiated pollen and mixtures of irradiated and self pollen. Consequently, 995 seedlings from 9,052 seeds were screened for 1,594 potentially detectable transformation events. A very low frequency of unexpected progeny resulted, but these results were not repeatable and appear to have arisen by mechanisms other than transformation. These results are strongly at odds with previous claims, when 50% of offspring were found to be transformed. We conclude that the previous observations of high frequency egg transformation are not reproducible. However, due to the ambiguous nature of the markers employed, it is very difficult to prove that such transformation events do not occur as very rare events.