This　study　optimizes　the　fleet　size　and　schedules　of　feeder　buses　that　connect　metro　and　residential　areas　in　the　context　of　bike-sharing　systems.　We　propose　hybrid　operation　modes　that　combine　fixed　and　dynamic　frequencies　in　a　bimodal　period　in　comparison　with　the　conventional　bus　scheduling　with　constant　service　frequencies.　The　effect　on　endogenous　demands　from　bike-sharing,　which　is　another　option　for　commuters,　is　considered.　This　study　proposes　a　multi-objective　model　under　one　hybrid　mode　with　morning　fixed　and　evening　demand-responsive　(MFED)　service　to　minimize　the　average　passenger　waiting　time　and　maximize　the　operator　profits.　The　constraints　include　vehicle　capacity　and　passenger　mode　choices.　Two　algorithms　are　developed　to　solve　the　feeder　bus　planning　problem　of　a　metro　station　and　three　nearby　communities　in　Chengdu,　China　(i.e.,　non-dominated　sorting　genetic　algorithm-II　[NSGA-II]　and　a　customized　multi-objective　optimization　algorithm　based　on　Particle　Swarm　Optimization　[MPSO]).　Numerical　results　show　that　the　proposed　MPSO　algorithm　slightly　outperforms　NSGA-II　in　terms　of　solution　quality　and　efficiency.　We　further　compare　two　feeder　transit　operation　modes　(i.e.,　morning　demand-responsive　and　evening　fixed　[MDEF]　service　and　all　fixed　service)　and　find　that　the　MFED　outperforms　MDEF　and　fixed　operation　mode　in　terms　of　system　effectiveness.　In　addition,　we　confirm　that　the　effect　of　bike-sharing　systems　cannot　be　neglected　because　passengers　may　change　their　travel　mode　while　waiting　for　a　feeder　transit　service.　This　study　can　provide　useful　policy　implications　and　operational　recommendations　for　government　agencies　and　transit　authorities　to　regulate　the　bike-sharing　market　for　effectively　addressing　the　first-and-last-mile　issue.　(C)　2019　Elsevier　Ltd.　All　rights　reserved.