In the creation and introduction of methods of monitoring and investigation of the properties and state of rock in situ there are two goals: minimization of costs of monitoring, and getting the maximum of objective information about the rock. Solution of these problems primarily involves optimization of methods of measurement in situ and corresponding interpretation of the results.One of the most important features of optimization of measurements is a well-founded choice of the information parameter for monitoring, which must be carried out with consideration of the main requirement: --detail, accuracy, and reliability of geomonltoring.In this article we discuss the comparative assessment of information parameters used in acoustic measurements of rock in place, and establish a relation between the index of Information content of the parameter | and the reliability of monitoring of the state of the rock.In geoacoustic investigations, the information parameters of monitoring are usually the following characteristics of the acoustic signal: a) Velocities of propagation of longitudinal Vp and transverse V s elastic waves. b) Amplitude A or damping factor u of elastic vibrations.To increase the effectiveness of geomonitoring, in a number of problems (e.g., in studying Jolnting in rocks) use is made of the spectral correlation characteristics of the acoustic signals (the displacement ~f/fo of the maximum of the spectral density towards lower frequencies, the correlation interval T of noise signals propagated in a fissured medium, etc.).The lack of an objective criterion for the information content of a control parameter has the result that the control parameter is chosen without sufficient consideration of the properties of the test object --the rock in place which exists in particular mininggeological and mining-technical conditions. Since a set of different factors {F} influences the characteristics of the acoustic signal propagated in the rock, including the structure of the rock and physical fields of various natures, not usually sufficiently well taken into account by experiments, we get a situation in which changes in the characteristics {| of the acoustic signals are random and the experimenter cannot uniquely determine which factor F(i) is causing an observed change in the information parameter | (Fig. i).Furthermore, by experimental investigations it has been established [i, 2] that one acoustic information parameter | is correlated with several factors F~, F2, ..., Fn, and conversely, one factor F is correlated with various acoustic parameters. Because these correlations are not equivalent, we must make'a quantitative estimate of the information content of the control parameters with allowance for inhomogenelty of the rock and the conMining Institute, Moscow.