Many failures in long-term collaboration occur because of a lack of activity awareness. Activity awareness is a broad concept that involves awareness of synchronous and asynchronous interactions over extended time periods. We describe a procedure to evaluate activity awareness and collaborative activities in a controlled setting. The activities used are modeled on real-world collaborations documented earlier in a field study. We developed an experimental method to study these activity awareness problems in the laboratory. Participants worked on a simulated long-term project in the laboratory over multiple experimental sessions with a confederate, who partially scripted activities and probes. We present evidence showing that this method represents a valid model of real collaboration, based on participants' active engagement, lively negotiation, and awareness difficulties. We found that having the ability to define, reproduce, and systematically manipulate collaborative situations allowed us to assess the effect of realistic conditions on activity awareness in remote collaboration.