Say no more than 5 publications.
Speaking of great mathematicians, I must name Harish Chandra (11 October 1923 – 16 October 1983). Well, he had more than 5 publications. but at most one or two per year. This Indian American mathematician worked on Lie groups.
I must add Grigori Yakovlevich Perelman (13 June 1966 – ) to this answer. This Russian mathematician has made landmark contributions to Riemannian geometry and geometric topology. Mostly famous for his work on Poincaré conjecture, this man was nominated for the Fields Medal in 2006, but declined to accept the award. This list shows an incomplete list of publications so far (but remember, he crossed 40 years!) - see 900 sit-ups a day's comment.
If ancient (but not so much) mathematicians are allowed, I propose
Maria Gaetana Agnesi
as far as I can check, she published three manuscript.
Oratio qua ostenditur artium liberalium studia a femineo sexu neutiqua abhorrere, Milano, In Curia Regia, 1727
Propositiones philosophicae, Milano, In Curia Regia, 1738
Instituzioni analitiche ad uso della gioventù italiana, Milano, Nella Regia-ducal Corte,1748
Yutaka Taniyama has only 5 papers in Math Reviews (and one of those was published 50 years after his death). He is the Taniyama of the Taniyama-Shimura-Weil conjecture, now known as the Modularity Theorem, which played a major role in the proof of Fermat's Last Theorem (now known as the Wiles-Taylor Theorem).
Depends on what you mean by great:
Urysohn has his lemma but I doubt he published much since he died really young.
As mentioned in the comments: Galois, Gauss etc (But if five is your criteria then Gauss will not count since he published different proofs of the fundamental theorem)
Probably Eisenstein of Eisenstein's criterion as he also died very young (He was Gauss's student and Gauss thought very highly of him).
Among Fields Medals, you have for instance Ngo Bau Chau who published about 15 articles up to his award, showing that it is not so far relevant to look at the number of publications.
Asking for if there are some with very weak H/G-index of publications would be an interesting question too... probably even more.
For more historical exemples beyond those already cited, there should be a lot, for institutional papers only emerged at the beginning of century XIX, and for long there weren't as much as in our recent times : mathematician without much credit or defending very original works did not succeed in publish :
- Cantor probably never would had published a lot without the backing of Mittag-Leffler
- Grassmann (with his original extension theory)
And, of course, Galois was far too young when he died for having published lots of things.
Moses Schönfinkel published a total of two papers. Still widely known as the father of combinatory logic. https://en.wikipedia.org/wiki/Moses_Sch%C3%B6nfinkel
Wolfgang Doeblin. I'm not sure whether he had any publications at all during his lifetime, but 60 years after he died someone got around to looking at some notes he had written and found Doeblin "was ahead of his time in the development of the theory of Markov processes. In recognition of his results, Itō's lemma is now occasionally referred to as the Itō–Doeblin Theorem."
Jacques Herbrand: I am not quite sure about the total of his publications, but as far as I know he has one important one in logic, and one in algebra.