Background/Aim: Non-tunneled, temporal hemodialysis (HD) catheters are commonly used as short-term vascular access for the HD procedure. One of their late complications is thrombotic occlusion of the catheter ensuing in their malfunction. A heparin lock is conventionally used for maintaining the patency of the catheter. The aim of the study was to evaluate the influence of heparin concentration used for locking the catheter canals (5,000 vs. 2,500 IU/ml) and some other clinical and laboratory variables at the time of temporal HD catheter functioning. Methods: Catheter malfunction was defined as the inability to attain and maintain a blood flow of at least 150 ml/min. 174 consecutive HD catheters inserted into jugular or femoral veins (114 patients) were followed up and remained in use for a total of 3,284 days. Results: Catheter thrombosis occurred in 53 cases (30.5%) during the study period, giving an overall rate of 16 episodes per 1,000 catheter-days at risk. In univariate Cox proportional hazard analysis, predictors of catheter dysfunction were: femoral localization (HR 4.92, 95% CI 4.30–5.50), acute renal failure (HR 1.75, 95% CI 1.18–2.32), higher mean ultrafiltration (UF) (HR 1.31, 95% CI 0.99–1.63) and higher concentration of hemoglobin (HR 1.15, 95% CI 0.99–1.33). The concentration of heparin used for canal locking did not influence the time of catheter functioning (HR 1.1, p = 0.7). In multivariate Cox proportional hazard analysis (χ2 = 38.5, d.f. = 4, p < 0.0001) the remaining statistically independent predictors of catheter malfunction were: femoral localization (HR 5.94, 95% CI 5.27–6.61, p < 0.0001) and higher UF (HR 1.60, 95% CI 1.24–1.94, p < 0.01). Conclusions: A lower concentration of heparin (2,500 IU/ml) prevents catheter thrombosis as effectively as a standard one (5,000 IU/ml). Femoral localization of HD catheters and higher UF during the HD procedure are the factors predisposing for catheter malfunction.