In an earlier paper we have shown, how one can successfully use constraint satisfaction techniques for proving and solving formulae in the first-order predicate language over the real numbers (i.e., real first-order constraints). This approach was restricted to inputs that contain inequality symbols such as ≤, but no equality symbols (=) or disequality symbols (≠). In this paper we lay the basis for extending this approach to inputs that contain (dis)equalities. This considerably widens the practical applicability of numerical constraint satisfaction methods.