Airborne particulate matter (PM) is studied because of its effects on human health and climate change. PM long-term characterisation allows identifying trends and evaluating the outcomes of environmental protection policies. This work is aimed to study the inter-annual variability of PM2.5 and PM10 concentrations and chemical composition in an urban background site (Italy). A dataset of daily PM2.5 and PM10 was collected in the period 2016–2017, including the content of OC, EC, major water-soluble ions, main metals, and compared to a similar dataset collected in the period 2013–2014. Oxidative potential using DTT assay (dithiothreitol) was evaluated and expressed in DTTV as 0.39 nmol/min·m3 in PM10 and 0.29 in PM2.5 nmol/min·m3. PM source apportionment was computed using the EPA PMF5.0 model and source contributions compared with those of a previous dataset collected between 2013 and 2014. Multi linear regression analysis identified which source contributed (p < 0.05) to the oxidative potential of each size fraction. Inter-annual trends were more evident on PM2.5 with reductions of biomass burning contribution and increases in traffic contribution in the 2016–2017 period. Crustal contributions were similar for the two periods, in both size fractions. Carbonates were comparable in PM10 with a slight increase in PM2.5. Sea spray decreased in PM10. The DTTV of PM2.5 peaked during cold periods, while, the DTTV of the PM10-2.5 fraction peaked in summer, suggesting that different sources, with different seasonality, influence OP in the PM2.5 and PM10-2.5 fractions. Analysis showed that sea spray, crustal, and carbonates sources contribute ∼13.6% to DTTV in PM2.5 and ∼62.4% to DTTV in PM10-2.5. Combustion sources (biomass burning and traffic) contribute to the majority of DTTV (50.6%) in PM2.5 and contribute for ∼26% to DTTV in PM10-2.5. Secondary nitrate contributes to DTTV in both fine and coarse fraction; secondary sulphate contribute to DTTV in PM2.5 with negligible contributions to DTTV in PM10-2.5. Display omitted •Source contributions have inter-annual variability higher for PM2.5 than for PM10.•Inter-annual variability of crustal, carbonates, and sea spray were lower than 0.3% of PM.•An increase of secondary nitrate and a decrease of sulphate in PM2.5 was observed.•DTTV of PM2.5 peaked in cold periods, instead DTTV in PM10-2.5 peaked in summer.•DTTV in PM2.5 and in coarse (PM10-2.5) fractions were influenced by different sources.