In this study, alternative methods for studying legibility of text while walking with a mobile phone were examined. Normal reading and pseudo-text search were used as visual tasks in four walking conditions. Visual performance and subjective evaluation of task difficulty were used as measures of text legibility. According to the results, visual performance suffers from increasing walking speed, and the effects are greater on reading velocity for pseudo-text search. Subjects also use more homogenous strategies when reading compared to pseudo-text search, and therefore it is concluded that reading is a more useful measure of legibility. Subjective measures are found to be more sensitive to small variations in legibility than objective measures, and give additional information about task demands. Hence, without both objective and subjective measurements important information about legibility in different conditions and with different tasks will be lost.