Background: Despite advancing technology for single-incision minimal access surgery, there have been no reports on the optimal operating setup and the instruments for use through a single port. This study therefore aimed to compare task performance in single-port surgery between crossed and uncrossed instruments (straight and distally coaxially curved). Methods: A dedicated trainer box was used to investigate the performance of two tasks (pick-transfer-place and dissection) by 18 surgical residents using two randomly allocated manipulating systems: (1) a conventional, uncrossed manipulation system (UCMS), with surgery through a cone-shaped 40-mm port using coaxially curved or straight instruments subtending a 30° manipulation angle, and (2) a crossed manipulation system (CMS), with surgery through a cylindrical miniport (25 mm) using coaxially curved instruments. Results: The pick-transfer-place task performed with the UCMS enabled comparison between straight and curved coaxial instruments and showed that although the participants moved the same number of objects with either type ( p = 0.464), they dropped a significantly more objects with straight instruments (1.5 ± 1.33 vs. 0.61 ± 0.85; p = 0.013). The execution of this task with the CMS (curved instruments only) provided a performance (dropped objects and error rates) equal to that obtained by the UCMS with curved instruments ( p = 0.521, p = 0.989). The dissection task with the UCMS showed no difference between straight and curved coaxial instruments in dissection accuracy (75.6 vs. 75.5 %; p = 0.950) or execution times (258 vs. 302 s; p = 0.367). The performance of this task by the CMS and the UCMS (with curved coaxial instruments) was equivalent with regard to dissection accuracy ( p = 0.849) and execution time ( p = 0.402). Conclusions: Using the UCMS, task performance is better with distally curved coaxial instruments. The CMS (with curved instruments) gives a task performance equivalent to that obtained with UCMS. [ABSTRACT FROM AUTHOR]