Background The increasing attention on functional assessments in medical and vocational rehabilitation requires a focus change for the general practitioners (GP) into paying attention to patient resources, possibilities and coping instead of symptoms, problems and limitations. The GPs report difficulties in performing the requested explicit functional assessments. The purpose of this study was to implement a structured method in general practice for assessing functional ability in persons with long-term sick leave. The study aim was to evaluate intervention effects on important GP parameters; knowledge, attitudes, self-efficacy towards functional assessments and knowledge about patient work factors. Methods Fifty-seven GPs were randomly assigned to an intervention or a control group. The intervention group GPs attended an introductory one-day work-shop and implemented structured functional assessments during an eight months intervention period. GP knowledge, GP attitudes, and GP self-efficacy towards functional assessments, as well as GP knowledge of patient work factors, were collected before, after and six months after the intervention period started. Evaluation score-sheets were filled in by both the intervention GPs and their patients immediately after the consultation to evaluate the GPs' knowledge of patient work factors. Results The intervention GPs reported increased knowledge (B: 0.56, 95% CI (0.19, 0.91)) and self-efficacy (B: 0.90, 95% CI (0.53, 1.26)) towards functional assessments, and increased knowledge about their patients' workplace (B: 0.75, 95% CI (0.35, 1.15)) and perceived stressors (B: 0.55, 95% CI (0.23, 0.88)) with lasting effects at the second follow-up. No intervention effect was seen in relation to GP attitudes. Both before and after the intervention, the GPs were most informed about physical stressors, and less about mental and work organisational stressors (Guttman's reproducibility coefficient: 0.95 and 1.00). After the consultation, both the intervention GPs and their patients reported that the GPs' knowledge about patient work factors had increased (GP B: 0.60 (95% CI: 0.42, 0.78); patient B: 0.50 (95% CI: 0.34, 0.66)). Conclusion Introducing and implementing structured functional assessments in general practice made the GPs capable to assess functional ability of their patients in a structured manner. Intervention effects of increased GP knowledge and GP self-efficacy sustained at the second follow-up.