A variety of biological processes can be studied from the washout of compounds labelled with a gamma emitter. Conventional systems for measurement of gamma radiation can detect count rates at both the low and high energy range of the indicator in question, 133Xe. However, silicon (Si) photodiodes have the capability of measuring the lowest energies (25-40 keV) of gamma and x-ray emitters with sufficient efficiency when applied on the skin surface and close to the indicator depot. The purpose of the present study was firstly to evaluate a portable Si photodiode matrix detector system, composed of 16 photodiodes, for in vivo measurement of disappearance rate constants of 133Xe following injection into subcutaneous (SC) or skeletal muscle (SM) tissues as compared with a stationary sodium iodide single crystal detector. Secondly, from these results, to evaluate the contribution from intra-tissue indicator diffusion and indicate means of avoiding influence from local diffusion in the washout curve of 133Xe. From the recorded 133Xe distribution in the depot proper, Si photodiodes in the matrix were selected for calculation of the washout. The Si matrix detector was reliable with a constant linearity and sensitivity within the range 20-38 degrees C and calibrated to +/- 1% between diodes. The in vivo investigations comprised two SC injections on humans with four measuring periods for comparison, and four SC and five SM injections on dogs, resulting in eight and five measuring periods for comparison. Only when the results obtained from the Si photodiodes at the position just above the count rate maximum were selected, was the contribution from diffusion significant. The ability to detect geometry changes is advantageous in studies with potential motion artefacts, such as during exercise.