Task efficiency and knot strength was evaluated under standardized conditions using direct vision and electronic imaging with two-dimensional (2-D) and three-dimensional (3-D) systems. Three operators with different endoscopic surgical experience tied a surgeon's knot with standard endoscopic instruments using the three different visual systems in random order. Each operator tied 20 knots with each visual system. Median task efficiency (defined as the time to complete the knot) was 35.0 (interquartile range (i.q.r.) 30.3-43.8) s for direct vision and 53.0 (i.q.r. 45.3-62.8) s and 53.5 (i.q.r. 45.0-64.8) s for 2-D and 3-D imaging respectively (P < 0.05). With respect to direct vision, this represented an overall degradation of task efficiency with the use of electronic imaging of 52 per cent, with no detectable difference between 2-D and 3-D imaging. The knot strength, representing the degree of tightening, was weaker with electronic imaging but the difference was not significant due largely to variation between the three operators.