The present study develops the given theme “Mathematics and Reality” along two lines. First, we explore the answers, in its various facets, to the following question: How is it possible that mathematics shows such wondrous efficiency when explaining nature\ulcorner In addition to a comparative analysis between empiricism and rationalism, constructivism as a function of idealism is compared with realism within the frame provided by rationalism. The second step involves limiting our discussion to realism. We attempt to explain the various stages of mathematical realism and their points of difficulty. Postulate of parallels, Godel's theorem, continuum hypothesis and choice axiom are typical examples used in demonstrating undecidable propositions. They clearly show that it is necessary to mitigate the mathematical realism which depends on bivalent logic based on an objective exterior world. Lowenheim-Skolem theorem, which states that reality is composed not of one block but rather of diverse domains, also reinforces this line of thought. As we can see the existence of undecidable propositions requires limiting the use of reductio ad absurdum proof which depends on the concept of excluded middle. Consequently, it becomes obvious that bivalent logic must inevitably cede to a trivalent logic since there are three values involved: true, false, and undecidable.