One hundred eighty four families completed a twelve month parent-guidance (experimental) or a parent telephone-monitoring (comparison) intervention initiated during one parent's terminal cancer illness and continued until six months after the death. Children in the parent-guidance intervention reported greater reduction in trait anxiety and greater improvement in their perceptions of the surviving parent's competence and communication, a primary goal of the intervention. Identified problems in implementing evaluations of experimental interventions with bereaved children include the following: (1) Available and commonly used standardized psychopathology measures do not adequately capture changes in non-psychopathological but bereaved distressed, grieving children and adolescents. (2) Experimental and control samples usually have very few children with psychopathology (scores). Relatively small changes in scores within the normal range may be insufficient to allow measurement of meaningful differences between interventions. (3) Both experimental and control interventions must provide sufficient help to retain families for later evaluation. The level of general support and referral for other treatments, if adequately done, may be sufficient to blur differences in standardized psychopathology measure scores between any two interventions. It may only be in the specifically targeted intervention area that differences can be expected to be significant in adequately resourced families.