A factorial matrix consisting of 25 cigarette samples, where tobacco density, circumference, paper porosity and paper citrate were varied at three levels was tested on cotton duck, treated sailcloth and four commercial fabrics. Duck fabrics are heavy canvas-type fabrics that have very high levels of potassium ions. Treated sailcloth is a light areal density fabric treated with potassium sulfate and hydroxide salts to produce ignition results that mimic those obtained with light areal density commercial fabrics. The chosen commercial fabrics had light, heavy and mid-range areal densities. The results underscore the fact that changing tobacco column parameters does not produce cigarette designs that perform consistently well on all fabric types. Lower cigarette density and reduced circumference levels reduce ignitions on duck fabrics but increase ignitions on treated sailcloth and the light areal density commercial fabric. Reducing paper porosity reduced ignitions on duck fabric, had little effect on treated sailcloth and increased ignitions on the light areal density commercial fabric. Paper citrate had the smallest effect. Ignition results on duck fabrics correlated with each other but were in contrast to those obtained with treated sailcloth. The heavy and light areal density fabrics produced similar ignition results to the duck fabrics and treated sailcloth, respectively. The two mid-weight commercial fabrics produced ignition results dissimilar to the other fabrics.