Objectives: To assess the effect of a pharmacist telephone counseling intervention on patients' medication adherence.Design: Pragmatic cluster randomized controlled trial.Setting: 53 Community pharmacies in The Netherlands.Participants: Patients ≥18 years initiating treatment with antidepressants, bisphosphonates, Renin-Angiotensin System (RAS)-inhibitors, or statins (lipid lowering drugs). Pharmacies in arm A provided the intervention for antidepressants and bisphosphonates and usual care for RAS-inhibitors and statins. Pharmacies in arm B provided the intervention for RAS-inhibitors and statins and usual care for antidepressants and bisphosphonates.Intervention: Intervention consisted of a telephone counseling intervention 7–21 days after the start of therapy. Counseling included assessment of practical and perceptual barriers and provision of information and motivation.Main outcome measure: Primary outcome was refill adherence measured over 1 year expressed as continuous outcome and dichotomous (refill rate≥80%). Secondary outcome was discontinuation within 1 year.Results: In the control arms 3627 patients were eligible and in the intervention arms 3094 patients. Of the latter, 1054 patients (34%) received the intervention. Intention to treat analysis showed no difference in adherence rates between the intervention and the usual care arm (74.7%, SD 37.5 respectively 74.5%, 37.9). More patients starting with RAS-inhibitors had a refill ratio ≥80% in the intervention arm compared to usual care (81.4 vs. 74.9% with odds ratio (OR) 1.43, 95%CI 1.11–1.99). Comparing patients with counseling to patients with usual care (per protocol analysis), adherence was statistically significant higher for patients starting with RAS-inhibitors, statins and bisphosphonates. Patients initiating antidepressants did not benefit from the intervention.Conclusions: Telephone counseling at start of therapy improved adherence in patients initiating RAS-inhibitors. The per protocol analysis indicated an improvement for lipid lowering drugs and bisphosphonates. No effect for on adherence in patients initiating antidepressants was found.The trial was registered at www.trialregister.nl under the identifier NTR3237.