Health professionals and patients tend to confuse drugs with similar names, thereby threatening patient safety. One out of four medication errors voluntarily reported in the US involves this type of drug name confusion. Cognitive psychology offers insight into how and why these errors occur. The objective of this investigation was to examine the effect of orthographic (i.e., spelling) and phonological (i.e., sound) similarity on the probability of making recognition memory errors (i.e., false recognitions). Prospective, computer-based, recognition memory experiments on 30 pharmacists and 66 college students were conducted. Participants viewed a study list of drug names and then a test list. The test list was twice as long as the study list and contained distractor names at progressively increasing levels of similarity to the study words. The task was to identify which test names were on study list and which were new. The main outcome measure was probability of making a false recognition error (i.e., of saying a new name was on the study list). Among pharmacists and college students, there was a strong and significant effect of similarity on the probability of making a false recognition error. It was concluded that both orthographic (i.e., spelling) and phonological (i.e., sound) similarity increase the probability that experts and novices will make false recognition errors when trying to remember drug names. Similarity is easily and cheaply measured, and therefore, steps should be taken to monitor and reduce similarity as a means of reducing the likelihood of drug name confusions.