The purpose of the monetary theory represented by the equation of Cash-Balance-or Transaction type is to explain the determination of the price level or money, prices. This monetary theory can give a complete explanation of pricing process in a monetary economy when it is taken together with the classical general equilibrium theory which explains the determination of relative prices. The monetary theory like this is really invalid. Why is it so? What monetary theory is valid? And what is the condition for constructing the valid monetary theory? It is the purpose of this paper to investigate these problems. We start with examining the classical system in order to know what role monetary theory plays in it. Consequently, we find out the following. The general system of equilibrium equations in the classical system is able to determine only relative prices. One more equation is needed to transform them into money prices. Monetary theory is considered as what provides it. Thus, an equation devised by such a monetary theory is Cash-Balance or Transaction equation. The property of this monetary theory, therefore, is called "complementary property". Monetary theory characterized with "complementary property" was criticized by J. R. Hicks and P. N. Rosenstein-Rodan in 1930's. Their criticisms were based upon a notion that uncertainty was the only factor which gave rise to demand for money to hold. According to this notion, there is no demand for money under static or stationary assumptions on which the classical system is set up. It is clear that monetary theory consistent with those assumptions can not give a meaningful equation. But A. W. Marget, soon after Hicks, proposed a problem as to possibility of static monetary theory, worked out a factor explaining demand for money in static economy, and it has-been recently named "time factor" by J. C. Gilbert. Time factor is a lack of synchronization between receipt and outlay. We, therefore, can assume non-zero money stock in a static model, too, only if it is developed to include the time assumption. But if we explicitly assume non-zero money stock in, the classical system, it becomes indeterminate. So money stock in it must be zero at all rates. This means that any complementary, monetary system spliced on the classical system is invalid. Criticisms from this viewpoint was made by D. Patinkin. According to him the only way to construct the valid monetary theory was to introduce money into every individual's utility function. It, however, has been made clear by K. Brunner that Patinkin was wrong. He has proved that it is possible to construct the monetary theory by relating money to the objects which enters utility function. The monetary theory thus obtained is not only a theory to give an equation for transforming relative prices into money prices. It is also a part of the theory of relative prices. Now, we can understand that in a system describing a monetary economy rightly, both relative prices and money prices should be directly determined by the general system of equilibrium equations. As a result of the above examination, we get the following conclusions. (1) Any valid monetary theory must get rid of "complementary property". (2) Any valid monetary theory proves that there exists a unique set of money prices in a monetary economy. (3) The conditions that Patinkin and Brunner put forth are really sufficient conditions. Whichever of them we may choose, it suffices as conditions for constructing valid monetary theory.