In 'beneficial inflammation', which is the major component of our innate immune system, it is possible to predict an 'ideal' sequence of cellular events: neutrophil migration would be rapid; time of contact with endothelial cells minimized; matrix degradation localized, with specific turn-on and turn-off of degradation mechanisms; neutrophil secretion and disintegration would be kept to a minimum during bacterial killing; and finally, rapid cessation of neutrophil migration and rapid removal of intact senescent cells would occur. Any doubts that the cellular events of the early stages of acute inflammation normally involve highly sophisticated cellular interactions, presumably designed to minimize tissue perturbation, should be dispelled by two elegant recent studies of neutrophil-endothelial interaction. Clearly, defects in the control of these processes could tip the balance towards cell injury or excessive matrix degradation and initiate amplification mechanisms leading to persistent inflammation and disease. The further identification of molecular mechanisms of these events should permit specific intervention in neutrophil-mediated disease. However, it is important to remember, firstly, that the neutrophil is just a part of the highly redundant inflammatory process and the removal of one 'strand' does not mean that the whole 'web' breaks down, and secondly, that impairment of neutrophil mechanisms may critically impair our anti-bacterial defences. Therefore, continued attempts should be made to define how cells and mediators interact in concert, to determine the fine specificity of molecular mechanisms and, in parallel, to identify 'time windows' in diseases, during which these mechanisms are more critical to the processes damaging the host than they are essential to its defences.