Three experiments provide evidence for a primacy effect in judgments of spatial location. Participants viewed and immediately estimated a series of spatial locations that were serially ordered from left to right or from right to left. In a subsequent block, they judged the rightmost, leftmost, and center of the distribution or were shown dots at those locations, which they then estimated from memory. Both judgments and memories were biased toward locations that had been presented earliest in the sequence. The findings indicate that participants incorporate not only geometric categories, but also aspects of their prior spatial experience, when estimating locations. The results mirror recent evidence for a primacy effect in nonspatial category induction, suggesting that this effect generalizes across domains.