In this study, the authors investigated the effectiveness of pictograms for steps in manufacturing procedures. Pictograms for manufacturing steps, such as "cut", "push" and "measure," were designed for the study. Their effectiveness was examined by means of both questionnaires and a comprehension experiment related to the subjects' response actions to the pictograms. The questionnaires were presented to 72 Japanese and 40 Portuguese-speaking subjects. The comprehension experiment was conducted with a different group of 69 Japanese subjects. Results from the questionnaires and the experiment suggest that the meaning of a pictogram containing an object can be comprehended better than one without an object. In particular, the pictograms for procedures performed by a body part, such as pushing by a finger and stepping by a foot, induced erroneous responses when they were presented without an object. On the other hand, the pictograms for procedures using a tool, such as cutting with scissors and measuring with a scale, tended to be comprehended more correctly and spontaneously than the procedural pictograms containing body parts.