A model of the scientific knowledge process is proposed in which the demands of practical applications have influence on what knowledge claims are accepted in science. However, this influence is limited in ways that prevent it from threatening the reliability of scientific knowledge. When practical applications put higher demands on reliability than the demands of knowledge per se, then the criteria for acceptance of the knowledge claims in question are adjusted upwards (epistemic adjustment). But in cases where practical decisions have to be based on information that is not reliable enough for the purposes of knowledge per se, the adjustment is made in the criteria for decision-making, not in those for acceptance of scientific knowledge claims (decisional adjustment). It is proposed that this model provides a more realistic account of the ideal that science should strive for than models requiring that the context of application have no influence at all on the acceptance or rejection of scientific knowledge claims.