In spinal cord injured patients with a reflex urinary bladder urodynamic evaluation of the detrusor pressure during the emptying phase is important, but the methods are not yet standardised. The aim of this study was to examine whether the detrusor pressure is significantly affected by the presence of a catheter in the urethra. In seven patients with a spinal reflex bladder, the maximum detrusor pressure and the duration of detrusor contractions in cystometry with a suprapubic technique were compared with the cor responding measurements when a closed 12F catheter was added to mimic a transurethral cystometric technique. Four cystometries were performed with about 10 min intervals and the mean values from two cystometries without the urethral catheter were compared with the mean values from two cystometries with the catheter. After addition of the urethral catheter there was an increase of the mean maximum detrusor pressure form 8,4 to 10.5 kPa (P = 0.009). The mean duration of the detrusor contraction increased from 122 to 191 s (P = 0.031) and the mean time during which the detrusor pressure exceeded 4 kPa, in each contraction, increased from 60 to 150 s (P =0.009). The average flow rate, calculated as the voided volume divided by the duration of a contraction, decreased from 1,4 to 0.6 ml s-1 (P = 0.009). Because of the observed obstructive effect, cystometric evaluation of the detrusor pressure during the emptying phase with a urethral 12F, or larger, catheter will in patients with a spinal reflex bladder, with a high probability, give a detrusor pressure of a higher amplitude and longer duration and a lower flow rate compared with the normal daily pathophysiological conditions, or cystometries with a suprapubic tech nique. This might not be a severe problem if for example some kind of therapeutic procedure is evaluated with the same technique before and after treatment. However, when comparing results of different types of studies, or in general discussions about pathophysiological mechanisms and risk levels, the observed difference has to be con sidered.