Small amounts (0.4 cc) of neutral water placed in small cylindrical cavities (5 mm diameter) in concrete exposed to 100% relative humidity first developed a pH comparable to that of a saturated solution. The pH then increased over a period of days-weeks toward a higher terminal value. A micro pH electrode arrangement was used. This behavior was observed in samples of 12 different concrete mix designs, including some with pozzolanic additions. The average terminal cavity pH closely approached that of expressed pore water from the same concretes. A simplified mathematical model reproduced the experimentally observed behavior. The model assumed inward diffusional transport of the pH-determining species in the surrounding concrete pore solution. The experimental results were consistent with the model predictions when using diffusion parameters on the order of those previously reported for alkali cations in concrete. The cavity size, cavity water content, and exposure to atmospheric should be minimized when attempting to obtain cavity pH values approaching those of the surrounding pore water.