Lamb primals (shoulders) were vacuum packaged or packaged in modified atmospheres containing 80% O(2)/20% CO(2), 50% CO(2)/50% N(2) or 100% CO(2), and stored at 5 or 0 °C. They were examined microbiologically at 7 day intervals for total counts obtained under (1) aerobic, (2) CO(2) enriched or (3) anaerobic conditions; B. thermosphacta; pseudomonad and Enterobacteriaceae counts. Off-odour assessments were also carried out at these times. In general, there were no significant differences between the total counts obtained from the different incubation conditions in any of the atmospheres. The only exception was noted in 80% O(2)/20% CO(2) at 5 °C. Significant differences between atmospheres for the total counts were observed at 0 °C only. In the case of B. thermosphacta, the pseudomonads and the Enterobacteriaceae, differences between atmospheres were noted at 5 and 0 °C. In general, vacuum packs and 80% O(2)/20% CO(2), and the two high CO(2) atmospheres fell into distinct groups. Storage temperature had a significant effect on all three counts. The relationship between bacterial counts and time was modelled using regression analysis. Data from total counts gave the equations of best fit. Significant differences between atmospheres in terms of off-odour production were observed at 5 °C only. The effect of temperature on off-odour production was significant in all atmospheres except 100% CO(2). A scheme was devised based on the growth of different groups of organisms which facilitated comparisons between studies on packaged meats. The results of the present work and that of others are discussed in relation to the different growth patterns which developed with packaging treatments and storage temperature.