Traditionally, it is believed that one of the most important phenomena in the history of "new" science, i.e. the science of Early Modern times, is the emergence of mathematical natural science. However, in the 16th century the status of physics and mathematics within the framework of scientific knowledge was far from being so unambiguous. In this article, we consider and analyze the arguments of the late Peripatetic author of the late 16th century – the learned Jesuit Benedict Pereira – in favor of his thesis about "non-scientific character" of mathematical disciplines. These arguments focus not on the weaker (less perfect) status of the reality of the mathematical object, but on the nature of mathematical demonstration and mathematical knowledge as such. Pereira shows in detail that mathematics does not meet the criteria of scientific knowledge (in the sense of "Second Analytics"), because the middle terms in its demonstrations are non-proper, general and accidental, and mathematics itself is not a knowledge of the real causes. In sum, in Pereira's consideration mathematics turns out to be some sort of “operational art” rather than a necessary knowledge of the truth from real causes. A comparison of the scientific status of physical and mathematical knowledge in Pereira makes it possible to clarify the conditions for the emergence of modern mathematical physics.