I study the effect of short-term interactions on scientific collaboration and knowledge flows. Using variation in the timing of when reviewers serve together as reviewers for the NIH, I find that short-term interactions lead to an increase in citations between pairs of scientists.