Hemolysis in clinical blood samples leads to inaccurate assay results and often to the need for repeated blood draws. In vitro experiments were conducted to determine the influence on hemolysis in phlebotomy needles and catheters of pressure difference, cannula diameter, and cannula material. Fresh blood from five human volunteers was forced from a syringe inside a pressurized chamber through 14, 18, and 22 gauge 304 stainless steel needles and polyurethane and Teflon catheters, all 40 mm long. Hemolysis was measured in the samples by a spectrophotometer. It was found that hemolysis increased with increases in pressure difference and cannula diameter and no consistent trend could be identified with regard to cannula material. The pressure differences required for significant hemolysis were above those typical of clinical venipuncture blood draws. While there was substantial variability among individuals, the hemolysis values scaled with exponent S = (t/t0)[(tau/tau0)-1]2, where t is the characteristic duration of shear, t0 is a time constant, tau is the wall shear stress, and tau0 is the wall shear stress threshold below which no hemolysis occurs. A hemolysis threshold including both time and shear stress was also defined for S = constant. The threshold implies that a threshold shear stress exists below which erythrocytes are not damaged for any length of exposure time, but that red cells may be damaged by an arbitrarily short period of exposure to sufficiently large shear stress.