The calcium transfection method for adenovirus DNA (Graham, F. L., and van der Eb, A. J. (1973). Virology 52, 456–467) has been improved by the use of 293 cells and a brief postinfection “glycerol boost.” Using this improved method, ts and hr mutations in intact DNA have been rescued with restriction endonuclease fragments of DNA bearing the corresponding wild-type genetic markers. For any one restriction enzyme, only one fragment rescues a given mutation, and fragments of as little as 4% of the total genome length are capable of rescue. The method therefore allows certain mutations to be mapped within very fine limits on the adenovirus DNA, and the coordinates of ts mutations mapped by restriction analysis of intertypic recombinant DNA (Sambrook, J., Williams, J. F., Sharp, P. A., and Grodzicker, T. (1975). J. Mol. Biol. 97, 369–390) have been confirmed and, in most cases, refined using marker rescue. A number of mutations have now been located unequivocally within the left quarter of the genome, and of these, the host-range mutant hr1 maps in the extreme left end. That location coincides with the “transforming region” of the adeno chromosome defined by other methods (Graham, F. L., Abrahams, P. J., Mulder, C., Heijneker, H. L., Warnaar, S. O., De Vries, F. A. J., Fiers, W., and van der Eb, A. J. (1974). Cold Spring Harbor Symp. Quant. Biol. 39, 637–650; Flint, J., Sambrook, J., Williams, J., and Sharp, P. A. (1976). Virology 72, 456–470) so it is of immense interest that we have found the hr mutants to be defective for transformation (Graham, F. L., Harrison, T., Williams, J. (1978). Virology 86, 10–21).