ABSTRACTCritical path analysis is proposed as a framework for problems of scientific productivity. Scientific productions are critical path problems because a single failed step in a project is sufficient to ruin an experiment or keep a paper from being published. For critical path processes, the probability of failure is a power function of the number of steps. Thus, project success is nonlinearly related to both the skill level and the number of sequential or critical path steps. A queuing model simulation of scientific production is presented. Results for short and long productions are shown to match documented lifetime publication rates of the most prolific (high skill level) and of average scientists (lower skill level). In this model creativity may decrease productivity by shifting the focus to more time consuming problems, increase productivity if the creativity resolves bottlenecks or eliminates problem steps, or have no effect if technique (e.g., observational skill) dominates the project's time budget. Sociological factors affecting productivity can also be integrated into the model. The model has the advantage that it can be tested against scientific diaries and by constructing time budgets and failure analyses of ongoing projects.