Abstract    Objective: The aim of the present study was to investigate if inclinometer estimates of arm posture during paper mill work differed between periods of different arm visibility. The aim was also to identify strategies for and difficulties in assessing body-visibility on video. Methods: The study consisted of two parts, where part 1 aimed to investigate the validity of observation-based assessment of arm posture during paper mill work. Study 2 was a survey among people with experience in assessing the posture of paper mill workers from video. Results: The total time for the right upper arm visibility indicates that it was on average visible in 54%, partly visible 15 %  and not visible 31%.