This report describes synthetic conclusions obtained from studies of electrotransformation made on 4 cell types: mammalian cells (CHO), bacteria (E. coli), yeast (S. Cerevisiae) and intact plant cells (maize), A similar multistep pathway is always observed in spite of the significant differences between cell types. Electrotransfer of DNA through the cell membrane and envelope is controlled by the amount of plasmids which is bound before the pulse, Electrotransformation seemed to be controlled by the electric parameters in a similar way for all species, even walled cells. Results on intact plant cells showed that the wall may prevent the transfer, Nevertheless, maize cells were electrotransformed with a simple pretreatment. DNA electrotransfer was obtained after a prepulse plasmolysis, which induced a contact between plasmid and membrane before the pulse, No transformation was detected when plasmid was added after the pulse. Electrooptics experiments on bacteria showed that a plasmid dependent orientation took place during the pulse along permeabilization. A key second step in gene transfer appeared to be a strong anchoring of DNA to the permeabilised membrane or envelope during the pulse. This is clearly associated to a direct field effect. Different electrophoretic effects may be involved. In the third step of electrotransformation, the plasmid slowly crossed the cell envelope by a cell dependent, but electrophoresis independent, process. For all species we studied, a delay was always observed between the pulse and the time of total protection of the plasmid against DNAse action. The fourth and final step of electrotransformation was dependent on plasmid functionality in the cell, Replication and gene expression should occur to detect the transfered activity. (C) 1997 Elsevier Science S.A.