In Europe, more than 307 000 premature deaths have been attributed to fine particle exposure (EEA, 2021). In particular, northern France is affected by relatively high particle matter (PM) concentrations, frequently exceeding the new annual PM2.5 concentration of 5 μg m-3 recommended by the WHO (EEA, 2021). Therefore, the study of aerosol particle sources is essential to improve air quality and to better develop effective mitigation policies. In this context, an Aerosol Chemical Speciation Monitor (ACSM) and an aethalometer (AE33) have been running quasi continuously since 2016 at a suburban site in Lille, known as the ATOLL (Atmospheric Observations in liLLe) platform, located on the University of Lille campus. The objective is to provide a comprehensive dataset of the chemical composition of submicron aerosol particles (PM1), and to investigate Organic Aerosol (OA) sources in the region through source apportionment analysis using Positive Matrix Factorization (PMF). Wind and back-trajectory analysis are also used to help identifying the geographical origins of advected aerosol particles, especially since Lille is at the crossroads of transnational pollution transport from Belgium, Germany, the Netherlands and the UK. Here, we present the first long-term analysis of non-refractory PM1 (NR-PM1) from October 2016 until December 2020. The average NR-PM1 concentration was 9.7 μg m-3, which was dominated by OA (45.4 %) followed by nitrate (31.4 %), ammonium (13.4 %) and sulfate (9.4 %). To study OA origins, we used the rolling PMF algorithm, analyzing one-month windows shifted by one-week step. Such approach combined with welldefined constraints has the advantage of capturing the potential temporal changes in the source profiles throughout the seasons and years. The OA PMF analysis yielded two primary OA factors: a traffic related hydrocarbon-like OA (HOA) and biomass burning OA (BBOA), and two oxygenated OA (OOA) factors. HOA showed a fairly constant contribution to OA throughout the seasons (with an average contribution of 12 %), while BBOA varied from 8 % to 18.5 %, with a peak in winter due to increased emissions from residential wood combustion. The OOA factors were distinguished between their less and more oxidized fractions (LO-OOA and MO-OOA, respectively). They contributed substantially to the total OA mass, with an average of 74 % (32 % and 42 % for LO-OOA and MO-OOA, respectively), indicating a strong participation of aged/secondary processes to local aerosol levels. A regional signature of those OOA factors is confirmed via wind and trajectory analysis, identifying the impact of polluted continental air masses from central Europe. During winter, the correlation between BBOA and LOOOA suggests the influence of aged biomass burning on the latter, identifying at least half of OA to be associated to wood combustion during wintertime. This was confirmed by complementary PMF analyses including inorganic and black carbon components. Overall, this study provides a thorough analysis of submicron aerosol sources at the recently established ATOLL platform, depicting complex interactions between regional traffic, wood burning, and agricultural activities in the highly urbanized North-Western European region. This work was supported by the French Ministry of Environment, through the CARA program. It is also part of the Labex CaPPA project (ANR-11-LABX-0005-01), and the CLIMIBIO project, both also funded by the Regional Council “Hauts-de-France” and the European Regional Development Fund (ERDF).