The aim of this study was to analyze pilots' visual scanning in a manual approach and landing scenario. Manual flying skills suffer from increasing use of automation. In addition, predominantly long-haul pilots with only a few opportunities to practice these skills experience this decline. Airline pilots representing different levels of practice (short-haul vs. long-haul) had to perform a manual raw data precision approach while their visual scanning was recorded by an eye-tracking device. The analysis of gaze patterns, which are based on predominant saccades, revealed one main group of saccades among long-haul pilots. In contrast, short-haul pilots showed more balanced scanning using two different groups of saccades. Short-haul pilots generally demonstrated better manual flight performance and within this group, one type of scan pattern was found to facilitate the manual landing task more. Long-haul pilots tend to utilize visual scanning behaviors that are inappropriate for the manual ILS landing task. This lack of skills needs to be addressed by providing specific training and more practice.