Intercity　freight　connections　have　reshaped　urban　networks　and　posed　significant　urban　planning　challenges.　Previous　studies　mainly　used　annual　statistical　data　or　online　orders,　which　are　criticized　for　the　poor　timeliness　and　data　representativeness.　This　study　presents　a　method　to　construct　freight　flow　networks　by　analyzing　large-sample　truck　trajectories　using　spatiotemporal　and　semantic　data　mining.　The　spatiotemporal　interaction　patterns　and　influencing　factors　of　intercity　freight　connections　in　China　are　then　investigated.　Key　findings　include:　(1)　The　proposed　method　enables　multidimensional　and　dynamic　investigations　by　constructing　timely　and　refined　freight　flow　networks.　(2)　The　network　structure　and　spatial　pattern　of　freight　connections　in　China　show　monthly　stability,　significant　spatial　heterogeneity,　symmetry,　and　a　typical　scale-free　property.　(3)　Spatial　interaction　patterns　are　related　to　freight　transportation　scale.　High-freight-volume　regions　form　complete　and　homogeneous　networks　with　activity　space　expanding,　while　low-freight-volume　regions　exhibit　heterogeneity　with　connections　focused　on　core　cities.　(4)　More　than　92　%　of　the　socioeconomic　factors　exhibit　significant　nonlinear　enhancement　when　interacting　with　distance.　The　nonlinear　interaction　between　economic　development　level　and　distance　is　the　most　influential,　explaining　23.7　%–85.7　%　of　the　connection　strength.　These　findings　facilitate　reasonable　urban　policy　formulation　and　regional　industrial　collaboration.　©　2024　Elsevier　Ltd