For a one-off question, one can use MathSciNet. Their full search page has an option called "Ref Author" which searches for the authors cited in papers. For instance, as of the 8th of April 2004, there are 244 references to Erdos and 201 for Wiles (and exactly 1 for the author).
If one is asking for a longer-term measure, then there is a number called the impact factor. This is defined by the ISI to be the number of citations to journal items over a certain period, divided by the number of journal items published over that period. This is designed to measure how many of the journal articles are interesting and relevant enough that other articles cite them.
The impact factors for various mathematical journals are given here. The Bulletin of the American Mathematical Society and Inventiones Mathematicae do well consistently over several timeframes, but there are other journals which also score highly, some of which might not be those which would come to mind.
Of course, trying to answer the question "Does the Journal of non-Riemannian hypersquares have a real impact on the field?" by using one number is not always going to succeed, but at least it gives a measure that could be used to compare journals in completely different fields.
Another, older, ranking of mathematical journals can be found here.
[This article was at least subconsciously inspired by this previous article on YMN.]
