In this research, the sustainability approach is used in the design of time-dependent routes for urban school transport from the mixed load, in this a comparison is made between the mixed load and the simple load, considering the social, environmental and economic aspects, to contrast the proposed hypothesis, which states that mixed loading in the presence of dependent times has scenarios in which it presents better sustainability performances, compared to simple loading. For the validation of the hypothesis, a mathematical model was formulated, in which in addition to cost, measures of social and environmental performance were proposed. The social focus was on reducing the risk associated with assigning children to stops, the risk on the road, and minimizing the time between the time of leaving home and the start time of classes at school, this time It includes the travel time and the time the child waits in school before the day begins. The environmental aspect materialized in the reduction of CO2 emissions. The formulated model includes the dependent times in the school bus routing problem, restrictions are activated or deactivated that allow addressing the problem with either mixed or single load. As a strategy to solve the multi-objective problem, epsilon constraints are used for small instances and for larger instances, a multi-objective algorithm is used, based on iterative local search with intelligent perturbations, which are derived from solving mathematical models that seek to preserve diversity. And the extremes in each new perturbation. As a main conclusion, it is found that the mixed load presents better results in sustainability compared to the simple load, in the scenarios in which there is a contrast difference in the dependent times between schools and the dependent times between stops. The specific scenarios in which this situation occurs are: when the dependent times between schools are low and the times between stops are medium and high; and when the dependent times between schools are high and the times between stops are low or medium. The above is explained to the extent that the benefits of doing a smaller routing in the pickups and extending the routing a little in the deliveries and vice versa can be consolidated. The main contribution of the present investigation is the consideration of the school bus routing problem from the sustainability perspective. On the other hand, the consideration of risk as a social dimension is proposed as another contribution. Dependent times are also introduced in the problem as a better approximation to SBRP in urban settings; and finally, an Intelligent Local Search, ILS, smart with multiple neighborhoods and perturbances derived from an optimization process.