To select an appropriate standard floor impact source to simulate real floor impacts, objective and subjective evaluations of the floor impact sounds were conducted in a box-frame-type structure with reinforced concrete slab floors. The sounds simulated in the test were those that would result from an adult walking barefoot, children running and jumping (represented by a heavy-weight impact source, such as a bang machine or an impact ball), as well as those of a person walking in high-heels or a lightweight object being dropped (represented by a tapping machine). Similarity tests between human-made impact sounds and standard heavy-weight impact sounds were performed. Sound quality (SQ) metrics were used to predict the results of the similarity tests. These results showed that the impact sound of an impact ball is more similar to a human-made impact sound than the sound of a bang machine. A multiple regression analysis showed that loudness and roughness are significant factors describing the results of similarity judgment among SQ metrics. Much of the data from the standard impact sources, measured in reinforced concrete floors with rigid floor coverings, have been collected. An empirical relationship was established to convert the impact pressure sound level from the bang machine or tapping machine to that from the impact ball. This study indicates that the use of an impact ball is reliable for simulating human impact sounds.