The effects of handle curvature on physiological and psychological cost in a static holding task, and the differences between curved and straight handles in a two handed lifting task, were investigated. In the static holding task experiments, the subjects consisted of five male and five female college age students. Boxes to be lifted were of weight 88 or 128 newtons and were 425 millimeters (mm) long by 215mm wide by 395mm high and had a detachable handle mounted 130mm above their upper surface. Handles were concave upwards with a radius of curvature of 50, 76, 102, or 203mm, or were straight. Each cut out handle was 25mm wide and 100mm long with 25mm diameter rounded ends, and was in either 3 or 13mm thick wood. Measures of heart rate and perceived exertion and discomfort were taken during trials in which subjects lifted boxes with 20 different combinations of weight, handle thickness and handle curvature. Lifts were for 25 seconds with rest periods for 95 seconds. No effects of handle radius of curvature were found. A handle of mild curvature, 102mm, was chosen for the study involving the two handed lifting task. The subjects were six professional manual materials handling workers. Boxes weighing 9 or 13 kilograms, and measuring 400mm square, with handles on each lateral surface were lifted between conveyors at floor and waist height for 3 minutes, and heart rate was measured during the lifting. Two each of symmetric and asymmetric handle positions were tested. Both straight and curved handles were found to provide excellent handle to hand fit, and were not significantly different from one another on any measure. Symmetric and asymmetric handle positions were better suited to the lifting of heavier boxes and light but bulky boxes, respectively.