We present an analytical model of concurrent engineering, where an upstream and a down-stream task are overlapped to minimize time-to-market. The gain from overlapping activities must be weighed against the delay from rework that results from proceeding in parallel based on preliminary information. Communication reduces the negative effect of rework at the expense of communication time. We derive the optimal levels of concurrency combined with communication, and we analyze how these two decisions interact in the presence of uncertainty and dependence. Uncertainty is modeled via the average rate of engineering changes, and its reduction via the change of the modification rate over time. In addition, we model dependence by the impact the modifications impose on the downstream task. The model yields three main results. First, we present a dynamic decision rule for determining the optimal meeting schedule. The optimal meeting frequency follows the frequency of engineering changes over time, and it increases with the levels of uncertainty and dependence. Second, we derive the optimal concurrency between activities when communication follows the optimal pattern described by our decision rule. Uncertainty and dependence make concurrency less attractive, reducing the optimal overlap. However, the speed of uncertainty reduction may increase or decrease optimal overlap. Third, choosing communication and concurrency separately prevents achieving the optimal time-to-market, resulting in a need for coordination.