To establish the influence of the peritoneal sampling technique on the measurement of fibrinolytic capacity. Clinical study. University hospital, Germany. 40 peritoneal biopsy specimens were taken from 10 patients who were having elective colorectal resections. Peritoneal biopsy specimens were taken either with a biopsy punch (n = 20) or manually with forceps and scissors (n = 20). Extent of agreement in fibrinolytic activities between specimens taken with biopsy punch and manually. Major endpoint-peritoneal tissue plasminogen activator (t-PA) activity. Minor endpoints-peritoneal tissue plasminogen activator concentration, and concentration and activity of plasminogen activator inhibitior type 1 (PAT-1). Intra-assay agreement and the extent of agreement between the groups were evaluated by the method of Bland and Altman. Correlation of repeated measurements of t-PA and PAI-1 concentrations and activities from the same sample using the same ELISA kit was high (r = 0.93-0.99, p < 0.01). t-PA activities and concentrations between the groups correlated poorly (r= 0.60 and 0.66, p < 0.01) while no correlation at all was seen for PAI-1 concentration and activity between the groups (r = 0.6 and 0.1, p = 0.2 and 0.9). The mean differences between the groups ranged from -27% to -4.8%. The sampling technique considerably affects the measurement of peritoneal fibrinolytic activity.