The evaluation in real-life settings of services for the follow-up and control of hypertensive patients is a complex intervention, which still needs analysis of the roles, tasks, and resources involved in the basic items: patient, healthcare professional, and the interaction between the two. To evaluate the impact of patient-general practitioner (GP) short-messages-based interaction, isolated from other items, on the degree of hypertension control in the follow-up of medium-to-low-risk patients in primary care, a randomized controlled trial has been performed: 38 GPs enrolled 285 hypertensive patients who recorded the results of self-blood-pressure (BP) monitoring, heart rate, and body weight, and completed an optional questionnaire in an identical manner over a six-month period. The telemedicine group (TmG) sent the data to a telemedicine-based system that enabled patient-GP interaction; the control group (CG) recorded the data on paper and could only deliver it to their GP personally in the routine visits. In the TmG, the results were better, but not significantly so, for: 1) degree of hypertension control, in terms of the percentage of uncontrolled hypertensives at the final visit (TmG versus CG: 31.7% versus 35.6%; p = 0.47); 2) reduction in hypertension during follow-up, comparing measurements (performed by a professional) at the initial and final visits of systolic BP (15.5 versus 11.9; p = 0.13) and diastolic BP (9.6 versus 4.4; p = 0.40); and 3) adherence to the protocol within compliance levels of interest in a real-life follow-up service: >>50% (84.8% versus 73.3%) and >>25% (92.4.8% versus 75.4%) ( p = 0.053). Other factors such as average values of self-measured systolic BP, diastolic BP and heart rate, acceptability of the protocol, and median number of consultations and hospital admissions were similar in both groups. Outcomes show that, taken alone, the patient-GP short-messages-based interaction has very little impact on the degree of hypertension control in patients with this profile. In complex interventions, to discriminate the impact of each of its components in isolation will enable us to design an efficient follow-up service, little demanding in terms of healthcare professional dedication, and optimized in other basic aspects.