Whether science can reach conclusions with substantial worldview import, such as whether supernatural beings exist or the universe is purposeful, is a significant but unsettled aspect of science. For instance, various scientists, philosophers, and educators have explored the implications of science for a theistic worldview, with opinions spanning the spectrum from positive to neutral to negative. To delineate a mainstream perspective on science, seven key characterizations or "pillars" of science are adopted from position papers from the world's largest scientific organization, the American Association for the Advancement of Science. Based on those pillars and an examination of scientific method, I argue that the presuppositions and reasoning of science can and should be worldview independent, but empirical and public evidence from the sciences and humanities can support conclusions that are worldview distinctive. I also critique several problematic perspectives: asserting that science can say nothing about worldviews and the opposite extreme of insisting that science decisively supports one particular worldview; weakening science so severely that it lacks truth claims; and burdening science with unnecessary presuppositions. Worldview-distinctive conclusions based on empirical evidence are suitable for individual convictions and public discussions, but not for institutional endorsements and scientific literacy requirements.