Cigarette design variables such as filter ventilation are capable of lowering all smoke constituent yields to give a lower tar yield product. To selectively remove a Hoffmann analyte the reduction of the Hoffmann analyte should be greater than the overall tar reduction. An experimental design approach (central composite design) to investigate the effects of filter pressure drop, cigarette paper permeability, filter ventilation and lamina tobacco type on mainstream Hoffmann analyte/NFDPM (tar) ratios under ISO machine-smoking conditions will be presented. The filter pressure drop range is between 40 to 120 mm water gauge, cigarette paper permeability from 10 to 100 CORESTA units, on-line laser filter ventilation between 0-70% and four lamina tobacco blend styles are Virginia, Burley (uncased), Oriental and a 1:1 mixture of Virginia/Burley. Different fractions of mainstream smoke can be represented by benzo(a)pyrene (particulate phase), pyridine (semi-volatile), acetaldehyde (vapour phase) and carbon monoxide (gas phase). Results from the Virginia blend indicate that benzo(a)pyrene/NFDPM ratio remains constant over the filter pressure drop, paper permeability and filter ventilation ranges used. The pyridine/NFDPM ratio decreases with increasing filter ventilation indicating that it is selectively removed by this cigarette design parameter. The acetaldehyde/NFDPM ratio increases with increasing filter pressure drop, which means that the NFDPM is selectively removed in preference to acetaldehyde. Carbon monoxide is selectively removed by increasing the filter ventilation whilst NFDPM is selectively removed in preference to carbon monoxide with increasing filter pressure drop. A more detailed examination of final results will be presented. This study illustrates how conventional filter and paper design variables can influence the machine yields of Hoffmann analytes in relation to tar when using lamina tobacco blends.