The influence of housing system on the initial bacterial contamination of the eggshell was studied. Two long-term experiments were performed. Bacterial eggshell contamination, as expressed by total count of aerobic and Gram-negative bacteria, was periodically analysed for eggs from a conventional cage, a furnished cage with nest boxes containing artificial turf or grids as nest-floor material and an aviary housing system. Results were log-transformed prior to statistical analyses. For both experiments no systematic differences were found between the conventional cage and furnished cage. The type of nest-floor material in the nest boxes of the furnished cages also did not systematically influence the bacterial contamination. A possible seasonal influence on contamination with a decrease in the winter period (up to > 0.5 log cfu/eggshell) of total count of aerobic and Gram-negative bacteria was observed in the first experiment. The contamination with total aerobic flora was higher (more than 1.0 log) on eggs from the aviary housing system compared to the conventional and the furnished cage systems. For Gram-negative bacteria this was not the case. During the entire period of both experiments, independent of housing system, shell contamination was not influenced by age of hens or period since placing the birds in the houses. For the total count of aerobic bacteria a restricted positive correlation (r2 = 0.66) was found between the concentration of total bacteria in the air of the poultry houses and initial shell contamination.