Objective: Despite beneficial effects on communication and process measures, client-centred practice has been shown to result in poor functional outcomes. To examine a potential explanation for poor functional outcomes, this paper aims to assess whether in client-centred therapy more time is spent on diagnostic consultation and less time on actual treatment compared to usual care. Method: A multicentre cluster randomised controlled trial was performed. Thirteen hospitals and rehabilitation centres, 29 therapists and 269 outpatients with multiple sclerosis participated. Measurements included an inventory of diagnostic and treatment goals, the number of sessions, therapy duration and therapy intensity. Results: In client-centred therapy, more sessions were used for diagnostic consultation (10.9% points difference, p = 0.030); the time needed to formulate the first treatment goal was longer (11.4 days difference, p = 0.041); there was a tendency towards more goals directed to diagnostic issues (0.69 goals difference, p = 0.056), spending more hours on indirect issues (1.16 h difference, p = 0.051) and towards a longer total therapy period (1.56 months difference, p = 0.058) than in usual care. Conclusion: Client-centred therapy resulted in more intensive diagnostic evaluation and less intensive treatment. This suggests that client-centred therapy should be adjusted towards a more proportional distribution of time devoted to diagnostic evaluation versus actual treatment.