The population ageing process, the foreseen labor force shortage caused by the lower birth rate and the risk of loss of knowledge due to the higher proportions of workers near to retirement, are the three challenges stressing organizations’ abilities to manage changes in the composition of the workforce. These shifts, coupled with the demand for knowledge and skills increasingly more complex and the rapid obsolescence of know-how that accompanies the development of new technologies, involve older workers in a deep way and require actions aimed at their training and employability needs. In order to keep older workers active longer, HRM practices should be designed to fit their own characteristics and to keep their skills up to date during all the course of their working lives. Lifelong learning enhances the productivity of workers over their careers and helps to implement policies aimed at extending work careers. Supplying training to employees not only increases their productivity, motivation and job satisfaction, but also boosts self-confidence and self-perception, with corresponding decreases in anxiety and sense of inadequacy (Becker, 1964; Maurer, 2001). However, there is evidence that older workers are often excluded from lifelong learning and training programs because of negative stereotypes such as, for example, that they are resistant to change, less motivated to learn or display worst performance as compared to their younger colleagues. These negative views, coupled with economic reasons linked to the perspectives of the human capital theory (Becker, 1964) for which the time available to enjoy a return on investment in learning for older workers is too short, lead to lower investment in training for these employees (Warwick Report, 2006). Given that HR professionals play an important role in managing people and can influence training and development practices, the focus of this research is on the HR professionals’ propensity to train older workers. There is in fact a lack of empirical data regarding the key variables influencing HR professionals' decisions to allow/deny access to training. The core research questions of this study are: What factors influence HR professionals in deciding whether to approve training proposals for older workers? What kind of training are they more likely to recommend for older employees and in which organizational contexts? To explore these issues three factorial surveys – or vignette studies – have been submitted to 66 Italian HR professionals. The factorial survey technique allows researchers to present hypothetical descriptions of situations or persons, while randomly manipulating selected key characteristics (Ganong & Coleman, 2006). Each respondent was asked to assess the desirability of providing three different types of training to older workers with different profiles in different organizational contexts. The first factorial survey was about training activities aimed at updating job skills, the second concerned training to enhance internal mobility and the third regarded training to facilitate external mobility. In this way, it was possible to verify whether the propensity to train older workers varied as a function of the type of training proposed. The research presents a number of hypothesis on the relationships between the characteristics of older workers and the HR professionals characteristics on the propensity to provide training for them. Furthermore, because of the hierarchical structure of the data, multilevel mixed-effects linear regression models have been applied in order to cope with sampling effects on the standard errors. The results show that the overall propensity to train older workers is quite low, reaching its peak when training is aimed at facilitating external mobility. HR professionals do not consider training as a tool for improving performance or productivity for less productive workers, but as a means of retaining workers who already perform well. In addition, HR professionals are more prone to train highly-skilled older workers and older workers with lower absenteeism rates. The probability that an older worker will receive training is further increased if companies are facing a labour force shortage. The age that marks the beginning of the decline in training investment is 50 years old, with a sharp drop after 60. As expected, the respondents’ own age and role strongly influenced the dependent variables in all three surveys, thus confirming the in-group bias hypothesis with regard to their age and the out-group bias with regard to their role. In all three surveys, older HR professionals and HR professionals with no managerial responsibilities rated the training proposals for older workers more positively. In the first factorial survey, contrary to what expected, the results show that HR professionals working in larger organizations are less prone to train older workers.