Two issues in the writing of multiple-choice test items were investigated: a comparison of three versus four options; and the use of the inclusive "none of these" option versus a content option. Subjects were 220 introductory psychology students, who were enrolled at a large southwestern university, responding to a final examination in psychology composed of 48 items. Each item was written in four versions: (1) four options without "none of these"; (2) four options with "none of these"; (3) three options without "neither of these"; and (4) three options with "neither of these". Item analysis and test analysis comparing the manipulated item versions were conducted using item response theory. The three-option format was found to be less difficult than the four-option format, and the use of the "none of these" option resulted in more difficult items. There was no difference in discrimination between three- and four-option formats, a possible argument in favor of a three-option format. However, the study did indicate observed differences in the reliability favoring the four-option format. (SLD)