This study investigates the morning commute problem with both household and individual travels, where the household travel is a shared ride of household (family) members. In particular, it considers the situation when a proportion of commuters have to drive their children to school first and then go to work (household travel). For household travel, departure time choice is a joint decision based on all household membersÕ preferences. Unlike the standard bottleneck model, the rush-hour dynamic traffic pattern with mixed travelers (household travelers and individual travelers) varies with the numbers of individual travelers and households, as well as the schedule difference between school and work. Given the numbers of individual travelers and households, we show that by appropriately coordinating the schedules of work and school, the traffic congestion at the highway bottleneck can be relieved, and hence the total travel cost can be reduced. This is because, departure/arrival of individual and household travels can be separated by schedule coordination. System performance under schedule coordination is quantified in terms of the relative proportions of the two classes of travelers and is compared with the extreme case when the same desired arrival time applies to both schooling and working. Furthermore, the efficiency of work and school schedule coordination in reducing travel cost is bounded. This efficiency is also compared with that at the system optimum where queuing is fully eliminated and schedule delay cost is minimized (achieved by a joint scheme of first-best pricing and schedule coordination).