Purpose: The purpose of the paper is to highlight the training factors that mostly affect trainees' perception of learning and training usefulness. Design/methodology/approach: A new research model is proposed exploring the relationships between a trainer's performance, training programme components, outcomes of the learning process and training usefulness. The validity of this model is tested empirically adopting the structural equation modelling approach, using data from 126 employees who have participated in different training programmes. Findings: All the factors that constitute a training programme (trainer performance, training environment and training goals, content, material, process) have been found to affect both learning outcomes and training usefulness. Research limitations/implications: The long-term results of a training programme are not examined. Practical implications: Although all the aspects and components of the training programmes should be carefully considered by the training programme designer, training process and trainer performance have the strongest direct and indirect effect (respectively), on both learning outcomes and training usefulness. Originality/value: In this study, all the factors that constitute a training programme are incorporated in a single model using data from small and medium-sized enterprises. Overall, the final model can explain 39 and 60 percent of the variance in learning and training usefulness, respectively. (Contains 5 tables and 2 figures.)