Four inoculation techniques were compared for initiation of growth on cake surfaces: spot, air cabinet, spray (atomizer), and talc addition methods. Molds were isolated from commercial cakes and were identified as Aspergillus sydowii, Aspergillus ochraceus, Penicillium funiculosum, and Eurotium herbariorum. Cake surfaces were inoculated with mold spores and incubated under three equilibrium relative humidity (ERH) levels: 97, 85, and 75%. Random contamination by spores in a ventilated air cabinet was the simplest method of inoculation, but standard deviations in the inoculation rates (20% on a relative scale) were almost twice those observed with the other methods. The spot method was the most reproducible. Cake samples inoculated in the air cabinet had colony counts 10 times lower than those obtained for potato dextrose agar plates at 97% ERH, which was not the case with the spray and talc methods. Growth of molds was much slower in the samples incubated in 75% relative humidity, with all methods. Colony counts were generally similar in systems adjusted at 85 to 97% ERH but were lower for samples incubated at 75% ERH. In comparisons of the shelf life estimates obtained by the various inoculation methods, a correlation coefficient (r²) of 0.70 was obtained between the spot method and the other methods of inoculation, while talc, air cabinet, and spray shelf life data were correlated better (r²  0.97). The spot method appeared to be the method of choice in consideration of ease of use, precision, and the ability to enable the study of the effects of the environment on mold-free shelf life as well as on the rate of growth of molds on cakes.