Interventions to increase physical activity (PA) among older community-dwelling adults may be enhanced by using multidimensional health risk assessment (HRA) as a basis for PA counselling. The study was conducted among nondisabled but mostly frail persons 65 years of age and older at an ambulatory geriatric clinic in Bucharest, Romania. From May to July 2014, 200 participants were randomly allocated to intervention and control groups. Intervention group participants completed an initial HRA questionnaire and then had monthly counselling sessions with a geriatrician over a period of six months that were aimed at increasing low or maintaining higher PA. Counselling also addressed the older persons concomitant health risks and problems. The primary outcome was PA at six months (November 2014 to February 2015) evaluated with the International Physical Activity Questionnaire. At baseline, PA levels were similar in intervention and control groups (median 1089.0, and 1053.0 MET [metabolic equivalent of task] minutes per week, interquartile ranges 606.0-1401.7, and 544.5-1512.7 MET minutes per week, respectively). Persons in the intervention group had an average of 11.2 concomitant health problems and risks (e.g., pain, depressive mood, hypertension). At six months, PA increased in the intervention group by a median of 180.0 MET minutes per week (95% confidence interval (CI) 43.4-316.6, p = 0.01) to 1248.8 MET minutes per week. In the control group, PA decreased by a median of 346.5 MET minutes per week (95% CI 178.4-514.6, p<0.001) to 693.0 MET minutes per week due to a seasonal effect, resulting in a difference of 420.0 MET minutes per week (95% CI 212.7-627.3, p< 0.001) between groups. The use of HRA to inform individualized PA counselling is a promising method for achieving improvements in PA, and ultimately health and longevity among large groups of community-dwelling older persons. International Standard Randomized Controlled Trial Number: ISRCTN11166046.