Pressurciflow studies were performed in 28 men with benign prostatic hypertrophy. Twcnty-three of the men were also studied postoperatively. Urethral function during micturition was quantified by the urethral pressureiflow relation, pder = p,,,, + L,, Qm, where pdr, is detrusor pressure, p, , is minimal opcning prcssure, Q is flow rate, and m and L, are parameters. Using this method to quantify urethral function, three urodynamically different types of obstruction can he defined. In the first of these, pm0 is elevated corresponding to Schafer's compressive obstruction. The second is a constrictive type of ohstruction in which m ? 413 and L, is elevated and the third is a low-compliant type of obstruction in which m 5 I and L,,, is elevated. The two latter types of obstruction may be combined with a compressive obstruction.The preoperative pressureiflow relations were mostly characterized by a very high pmo, a moderately elevated L,,,, and a low m. Thus the majority of patients had a corn!% nation of compressive and low-compliant obstruction. Postopcratively, micturitions were much improved and the pressureiflow relations often had an even lower p , , > than normal, a normal L,,,, and a high m.From the pressureiflow relations, the clasticity of the flow-controlling zone can be estimated and described by the urethral prcssureiarea relation, p(A) = p,,,, + K, A", where p(A) is the intrinsic urethral pressure, A is the cross-scctional arca of thc flow-controlling zone and K, and n are parameters descrihing the distensibility of the flow-controlling zonc. Preoperatively, the flow-controlling zone had a low distensibility. The shape of the CUNC suggested that the urethra could have been distended further by higher pressure. Postoperatively. the urethra was distended to larger cross-sectional areas, but in many cascs thc shape of the curve suggested that distension was restricted by fibrosis. Median and rangc values for the model parameters as well as discriminating limits bctwccn the prcoperativc micturitions and micturitions in elderly men without voiding problems are presented. The exponent m is not perfectly reproducible but tends to be the same if a person performs several micturitions. The residual sum of squares is often increased more than 100% if an exponent m value other than the optimal one is used for curve fitting.Results when the micturitions were analysed with Schafer's model and classified using