As the cost of computers declines relative to outfitting and maintaining laser spectroscopy laboratories, computers will account for an increasing proportion of the research conducted in fundamental combustion science. W.C. Gardiner foresaw that progress will be limited by the ability to understand the implications of what has been computed and to draw inferences about the elementary components of the combustion models. Yet the diagnostics that are routinely applied to computer experiments have changed little from the sensitivity analyses included with the original chemkin software distribution. This paper describes some diagnostics capabilities that may be found on the virtual combustion science workbench of the future. These diagnostics are illustrated by some new results concerning which of the hydrogen/oxygen chain branching reactions actually occur in flames, the increased formation of NOx in wrinkled flames versus flat flames, and the adequacy of theoretical predictions of the effects of stretch. Several areas are identified where work is needed, including the areas of combustion chemistry and laser diagnostics, to make the virtual laboratory a reality.