The experimental and data processing methods used were improved in several respects compared with those described in previous publications. One hundred twenty-two stress decrease curves for bladder wall strips were measured at constant elongation; reproducibility was tested. Consideration of the results led to the addition of two elements (a dashpot and an active element) to our model of the process involved. The influence of the dashpot on the measurements is mathematically analyzed in terms of an increase in rest length during stretch. The model could be successfully fitted to the experimental data and the parameters turned out to be in the physiologic range. Some recent models used to describe muscular stress relaxation were tested and criticized.