Objective: To investigate the efficacy of exercise interventions and factors associated with changes in work ability for people with chronic whiplash-associated disorders. Design: Secondary analysis of a single-blind, randomized multi-centre controlled trial. Setting: Interventions were conducted in Swedish primary care settings. Patients: A total of 165 individuals with chronic whiplash-associated disorders grade II–III. Methods: Participants were randomly allocated to neck-specific exercise, neck-specific exercise with a behavioural approach, or prescribed physical activity interventions. Work ability was evaluated with the Work Ability Index at baseline, 3, 6 and 12 months. Results: The neck-specific exercise with a behavioural approach intervention significantly improved work ability compared with the prescribed physical activity intervention (3 months, p = 0.03; 6 months, p = 0.01; 12 months, p = 0.01), and neck-specific exercise at 12 months (p = 0.01). Neck-specific exercise was better than the prescribed physical activity intervention at 6 months (p = 0.05). An increase in work ability from baseline to one year for the neck-specific exercise with a behavioural approach group (p < 0.01) was the only significant within-group difference. Higher self-rated physical demands at work, greater disability, greater depression and poorer financial situation were associated with poorer work ability (p < 0.01). Conclusion: This study found that neck-specific exercise with a behavioural approach intervention was better at improving self-reported work ability than neck-specific exercise or prescribed physical activity. Improvement in work ability is associated with a variety of factors.