Although reliable data concerning toys' safety, durability, and appeal to children would be useful to parents, educators, and toy manufacturers, systematic procedures for evaluating toys and collecting such data have not been readily available. This paper describes a technology for carrying out such evaluations with children in free-play settings. Children entering the toy-evaluation area were each given a large tag, coded to indicate their names, age, and sex, to wear on the back of their shirt. Twenty-five toys were initially displayed on a table. Children were required by an adult to check out, and later return, any toy they selected for play. Every 5 min, an observer wrote down the tag number of any child playing with each of the 25 toys. The area was open for several hours daily, and at the end of each day a supervisor completed a toy-use questionnaire that identified problems associated with toy management, safety, and durability. Each interval in which a child was recorded as playing with a toy was considered five child-minutes of use for that toy. Data were summarized after each 5- to 10-day toy evaluation to indicate the total number of child-minutes of use attributed to each of the toys tested. The toys were then ranked from the most- to least-used. This measure of toys' overall use proved to be a reliable predictor of toys' appeal across many subjects and settings, and correlated well with our other analyses of toys' use. For example, toys used by individual children for long periods of time, and those toys most often chosen from the toy table were frequently the same toys that ranked high in overall child-minutes of use. Reliable differences in toys' appeal to boys and girls of different ages were observed. To test the generality of the toy-evaluation procedures, several experiments were conducted. To find out whether the popularity of certain toys depended on the number of children playing, a series of popular toys was made available first to only a single child at a time, then to any number of children. While certain toys required the presence of two or more children before they would be used, most toys were used by children both alone and in groups. The popularity of most toys, even the least-used toys, could be increased by having an adult play alone with the children. With some toys, an adult could maintain the play of several children at a time; with other toys, only one child at a time would be engaged. Another study showed a prolonged period of enforced sampling with little-used toys almost never increased their subsequent use once the enforced sampling procedures were discontinued. Finally, we found that having multiple copies of the same popular toys did not produce much more use of the toys than having single copies. The data produced by these evaluation procedures could be helpful to parents and educators in selecting desirable toys, and to manufacturers in improving the toys they develop and market. Future research might involve developing procedures that would encourage concerned adults to carry out toy evaluations in their own communities.