Computer users online encounter a variety of alerts, confirmation notices, and advertisements that have the potential to interrupt users in pursuit of their goals online. This dissertation examines cognitive and affective effects of interruptions, and it attempts to add to the existing body of literature on interruptions by varying the complexity of interruptions, timing of interruptions, and complexity of the primary task- which is reading online news stories. The current studies operationalized interruptions as pop-up advertisements that users had to contend with while trying to read news stories on a news Web site. Memory and attitude toward interruptions, stories, and Web site were measured in two studies. The first was a 2x3 between-subjects design that manipulated structural complexity of the interruption (simple and complex) and interruption timing (beginning, middle, and end) (N = 106). The experiment showed no main effects of structural complexity of an interruption or interruption timing on memory for interruptions, memory for stories, time spent on site, attitude toward interruptions, or attitude toward the Web site. However, participants' memory for interruptions decreased for complex interruptions that occurred later and increased for simple interruptions that interruptions occurred later. The second study was a 2 (structural complexity of the interruption -- simple or complex) x 3 (interruption timing -- beginning, middle, or end) x 2 (story complexity -- simple or complex) between-subjects design (N = 214). There were no main effects of structural complexity of an interruption or interruption timing on memory for interruptions, memory for stories, time spent on site, attitude toward interruptions, or attitude toward the Web site. Story complexity also exerted no main effects except for the amount of time spent on the Web site. However, when participants encountered simple interruptions, memory decreased for later interruptions in short stories. Conversely, memory increased for later interruptions in long stories. In addition, when participants encountered complex interruptions, memory increased as users encountered interruptions occurred later in short stories. In long stories, participants' memory scores were lower when interrupted in the middle than at the beginning or end. Theoretical and practical implications, limitations, and suggestions for future research are discussed.