The aim of this thesis is to define, model and solve three research questions at the tactical and operational levels of decision making, the former arising in the context of intermodal service network design and the latter two in passenger transportation, by using integer programming, dynamic programming and heuristics. The first research question concerns intermodal freight transportation, which is concerned with the shipment of commodities from their origin to destination using combinations of transport modes. Traditional logistics models have concentrated on minimising transportation costs by appropriately determining the service network and the transportation routing. The first chapter considers an intermodal transportation problem with a detailed consideration of greenhouse gas emissions and intermodal transfers. Two mathematical models, one time-invariant and the other time-dependent, are described for the problem, which are both in the form of a non-linear integer programming formulation, but which are linearised. A hypothetical but realistic case study of the UK forms the test instances for our investigation, where uni-modal with multimodal transportation options are compared using a range of fixed costs. The second and third research questions concern the multiobjective shortest path problem (MSPP) and the constrained multiobjective shortest path problem (CMSPP), extensions of the classical shortest path problem, with a wide range of practical applications particularly in passenger transportation. The second and third research questions are studied in two different chapters. The first of these presents several labelling algorithms for the MSPP and the CMSPP. Extensive testing is performed on different types of networks, including randomly generated and grid networks. The results show that label correcting algorithms are more efficient than label setting algorithms for solving both the MSPP and the CMSPP. The second of these two chapters proposes two fast local search algorithms for the MSPP. Four performance indicators are used to evaluate the local search solutions. Computational results demonstrate that local search algorithms are faster than all heuristic methods for the MSPP presented in literature, and able to produce reasonably good-quality solutions.