This study aims to investigate the effect of slider design and length on user performance and preference of smartphone versions of Visual Analogue Scale (VAS). Twenty-eight participants performed a task to set random target values with 8 smartphone versions of VAS: 2 slider designs (traditional design, modern design) × 4 slider lengths (4.3 cm, 5.8 cm, 10 cm landscape, 10 cm portrait). Experimental results showed that both slider design and length significantly affected the accuracy, task completion time and subject preference. Compared with the traditional slider design, the modern slider design showed significantly smaller bias in setting values, shorter task completion time, and higher subject preference. The slider length significantly affected all measures, and 5.8 cm was recommended due to small bias, short task completion time, dominant preference and excellent ability to closely fit the width of smartphone display with the portrait mode. These findings could provide mobile VAS and slider designers with useful references.