How travellers perceive the quality of public transport services has profound influence on their satisfaction and thus of their inclination to choose this mode of transport. Punctuality, which may be interpreted as timetable adherence or overall travel time predictability, is often pointed out as the most important constituent of overall service reliability of public transport systems. However, in order for operators to be able to improve passengers’ perceptions of service reliability, an empirically derived linkage between perceptions and actual service performance indicators must be established. This paper has addressed this issue by using six-year longitudinal dataset on half-year resolution level from the regional public transport system of Scania, Sweden, to relate various indices of actual service reliability, obtained from service performance indicators, to stated satisfaction with punctuality for trains and buses. This was done by means of linear regression models using a stepwise approach to eliminate redundant variables, based on the level of independence of each variable. The results indicate that average arrival delay and frequency of early departures in relation to schedule are the most important factors of service reliability to explain travellerś perceptions of punctuality. Thus, for instance, average arrival delay and schedule adherence are indicated to have three times larger impact on satisfaction than frequency of early departures. But we also find that other service quality factors such as cleanliness, safety, and seat availability are also indicated to have an impact on perceptions of service reliability. It is emphasized that appropriate service performance data, and their relation to perceptions of passengers, constitutes crucial information for commercial and public providers in their ambition to improve levels of passenger satisfaction with operations but that the optimal level of punctuality is likely to differ between different passenger categories, and thus services.