Modelling the mesoscale (10 to 100 years and 10 to 100 km) evolution of managed sandy shorelines is becoming increasingly necessary to guide the management of sandy coastal systems. Models that simulate mesoscale shoreline evolution assume an equilibrium active coastal profile. An equilibrium active coastal profile implies a fixed closure depth, defined as the seaward extent of the active coastal area, and shore-parallel depth contours, which present two limitations. First, an inability to account for sea-level rise, which will likely change the closure depth and be endogenous in coastal evolution over meso timescales. Second, an inability to account for complex planform morphologies where the closure depth varies longshore and depth contours are non-parallel. Such morphologies characterise sandy coastal systems in many vulnerable Caribbean islands where shoreline evolution models are most needed to guide coastal management. Hence, this thesis aims to create a method that accounts for sea-level rise and complex planform morphologies in mesoscale shoreline evolution predictions. Using a managed sandy coastal system in New York, Puerto Rico, and Southern California as test sites, I first assess the sensitivity of two mesoscale shoreline evolution models, MIKE21 and the Bruun Rule, to identify the most essential boundary conditions influencing shoreline evolution predictions in different coastal morphologies. I use the results of this sensitivity study to inform the development and application of three shoreline evolution modelling approaches, which include introducing: (a) a time-varying closure depth in MIKE21 as a solution to incorporate sea-level rise effects in mesoscale shoreline evolution predictions; (b) a space-varying closure depth in MIKE21 as a solution to account for complex planform morphologies in shoreline evolution predictions; and (c) a time and space-varying closure depth in MIKE21 as a solution to incorporate the effects of both sea-level rise and complex planform morphologies in meso timescale shoreline evolution predictions. Model sensitivity results show that nearshore discretisation, bathymetry, tides, friction and sediment properties are the key boundary conditions that influence shoreline evolution predictions regardless of the underlying morphology. I find that the optimal specifications of these boundary conditions match coastal system features, both morphology and processes. Specifying a time-varying closure depth in MIKE21 is found to provide a better alternative to the Bruun Rule for simulating mesoscale shoreline evolution under relative sea-level rise. However, I find that a time-varying closure depth causes MIKE21 to overpredict erosion over meso timescales, attributed to mean closure depth overestimation. Hence, there is a chance that a time-varying closure depth may improve mesoscale shoreline evolution predictions if closure depth time series estimations can be accurately prescribed. Enabling a space-varying closure depth in MIKE21 is found to replicate observed shoreline change in the Puerto Rico test site's complex planform morphology more realistically than existing modelling approaches. Lastly, allowing a time and space-varying closure depth in MIKE21 is found to provide theoretically plausible meso timescale shoreline evolution predictions under relative sea-level rise in the Puerto Rico test site's complex planform morphology compared to current modelling approaches.