Concurrent and retrospective verbal protocol methods were used to collect thoughts from 18 participants during a manual handling task involving the repeated transfer of loads between locations at two tables. The effectiveness of qualitative and quantitative methods of analysing the reported information was tested in the study. A simple taxonomy was developed to investigate the content of the reports (including reports on postures and loads) and determine how the participants approached the task (whether they made plans, described actions or evaluated their completion of the task). References to posture were obtained in the verbal protocol reports, indicating that the participants had some awareness of their postures during parts of the task. There were similarities in the content of the concurrent and retrospective reports, but there were differences in the amount of detail between the methods and differences in the way the reports were constructed. There could be some scope for developing the quantitative analysis of the frequencies of references to classes of information, though this can only be recommended for concurrent reports on tasks of short duration. The analyses of qualitative data gave a deeper insight into the reports, such as identifying factors that can be important when planning to handle a load, or illustrating how participants can change their focus of attention periodically throughout the task. The relative strengths of the concurrent and retrospective methods are described, along with ideas for improving the quality of information collected in future studies. A number of potential problems with the interpretation of the reported information are explained.