69 Background: In the Netherlands, many patients that might profit of palliative care don't receive it or don't receive it until the terminal phase. GPs in the Netherlands consider palliative care as an important task, but they don't feel confident to recognize the patients in need of palliative care and to provide it in a structured way. For that reason, we performed an RCT in which GPs in the intervention condition received training in recognozing the palliative patient and in early palliative care provision. Methods: Randomized controlled cluster trial (intention to treat) with GP as cluster. GPs in the intervention group received a 5-hours training in early recognition of the palliative patient, with the help of the RADPAC tool, and in proactive palliative care planning. They were invited to identify the patients in need of palliative care and provide the care as learned in the training. Primary outcome: number of out-of-hours contacts with the out-of-hours service, as measured retrospectively by analysing all patients that died during the intervention period of one year. Secondary outcomes: number of hospitalizations and place of death, also collected retrospectively. Results: Within one month, 159 GPs showed their interest to participate. Per responding GP, a mean number of 7 patients died during the intervention period. About 45% of the patients died at home, in both study groups. We did not find differences in out-of-hours contacts, hospitalizations and place of death between intervention and control group. Conclusions: We did not find differences between intervention and control group. As learning to identify the palliative patients was part of the intervention, we were not able to collect patient-related outcomes in the control group. This might explain why we did not find significant effects of early palliative care, in contrast to findings of Temel et al, and Zimmermann et al. and Bakitas et al. For that reason, we recommend to measure quality of life as primary outcome. Clinical trial information: NTR2815.