The aim of this paper is to identify the controls on the rates of illitization in Texas Gulf Coast shales under diagenetic conditions. Two different mineral transformations are considered: the transformation of smectite to (nearly) illite and the precipitation of discrete illite, i.e., the formation of pure illite. Analyses of potassium feldspar and kaolinite abundance, in addition to K 2 O, Al 2 O 3 , and TiO 2 contents, in shales from three wells are used to supplement published data in order to identify the mineral reactions responsible for the smectite-to-illite transformation in shales involving illite/smectite (I/S) mineral series. These data indicate that the first stage of the smectite-to-illite reaction in shales (R = 0 structures, ca. 0-50% illite content) is controlled by the presence of feldspar and hence also by the dissolution kinetics of feldspar. The empirically derived activation energy for the first I/S transformation in shales is similar to that for K-feldspar dissolution. Kaolinite is not involved in this transformation in the shales investigated. The second stage of the reaction sequence (R = 1 structures, ca. 50-100% illite content) occurs after potassium feldspar has largely or completely disappeared from the shale system. This indicates that the feldspar no longer controls the second stage of the smectite-to-illite transformation, and suggests another, external, source of potassium for the reaction in the shales. This assumption is supported by the relative increase of potassium content of the shales, at least in two of the three studied wells. Kaolinite does not seem to be involved either in this second stage of the I/S reaction sequence. The formation of discrete illite crystallites in shales is independent of the I/S transformation reactions and takes place only after a very significant period of time has elapsed. The slow rate of this reaction reflects the low activity of potassium in solutions that are in equilibrium with shale, in comparison with sandstones, where a rapid precipitation of pure illite can in some cases be observed. It is suggested that the different reactions in the smectite-to-illite transformation necessitate different sources of potassium either from within or outside an argillaceous whole-rock system. These different sources of potassium give rise to different rates for the I/S conversion reaction and the formation of illite.