In this study, three receptor models [factor analysis/multiple linear regression (FA/MLR), positive matrix factorization (PMF) and UNMIX] were applied seasonally to investigate the source apportionment of PAHs in the atmospheric environment of Alexandria, and a lifetime cancer risk was assessed. ∑44 (gas+particle) PAH concentrations varied from 330 to 1770ngm−3 and 170–1290ngm−3 in the summer and winter seasons respectively. PAH concentrations at the industrial sites were significantly higher than at the traffic and residential sites during the winter season (p<0.001). Summer PAH concentrations were significantly higher than the winter season at the traffic sites (p=0.027). Results obtained from the three receptor models were comparable. Vehicle emissions, both diesel and gasoline contributed on average 36.0–49.0% and 19.0–34.0% respectively, natural gas combustion 11.0–27.0% and, during the summer only, also evaporative/uncombusted petroleum sources 8.00–18.0%. Seasonal trends were found for the gasoline emission source. Overall, PMF and UNMIX models afforded better source identification than did FA/MLR. The lifetime cancer risk assessment showed that incremental lifetime cancer risks (ILTCRs) were greater than the acceptable level of 10−6 through dermal and ingestion routes at all the investigated sites and through the inhalation route at the industrial and traffic sites only. Total ILTCRs (6.64×10−3–4.42×10−2) indicated high potential risks to the local residents.