Warranty obligations in product service involve additional costs to the manufacturers. Building diagnostic features into a product can reduce the warranty costs because of; 1) fewer visits needed by warranty service representatives (through self-guided user repairs for simple problems) and 2) shorter service times (via reduced diagnostic time) for each failure. However, this approach increases the cost of the product and is worthwhile only if the reduction in the expected warranty cost is more than the additional costs incurred. This paper develops models to evaluate diagnostic design decisions for products under warranty from the manufacturer's point of view. The roles and influence of the key parameters and decision variables based upon the consideration of uncertainty are explored. Numerical examples are given to illustrate the proposed models.