Why are you matching and what is the exposure of interest? I ask because many dislike PS matching compared to PS weights, see research by Gary King as well as Peter Austin.
Not a Stata user, but would imagine it is about differences between score values (e.g., distance, caliper, or something).