I am an American High School teacher with a one-trimester course where the students are learning Esperanto. We have experimented this term with using Memrise for learning the vocabulary of our textbook. The purpose of this posting is to discuss our experience. We are six weeks into our course and I told students that they would earn an A at midterm for the vocabulary portion of the course if they put more than 500 words in their "long-term" memory (harvested to the garden). A large majority of students were able to do so, although many crammed to get it in during the last few days rather than pacing themselves as we covered the textbook. While not optimal, since it is midterm and student will need "water their garden" to keep the words active, I don't see this as a problem in terms of course goals. Students report liking the software and they feel that it keeps them engaged. Most have used some type of "flashcard" programs as well as other methods of vocabulary learning and so they are good sources of reference. And because I have a couple dozen students I can see how it works across a range of learning styles and abilities. Most students also think that the software algorithm can be made even better. The few student who were not able to get to 500 words was not because of aptitude but simply because they put off the task for too long and did not have the energy to put into the activity. Again, since it is only midterm, they will have more time to get with the program. While we gave some class time to it, most of the memrise activity was homework. I would estimate that students spent between 10 to 20 hours to master 500 words, although it would be great if the software was able to report actual time on task so that this would not need to be an estimate and so variation among users could be better understood. Reporting to teachers is a tricky question. More information on student progress could help the teacher tune his or her instruction. On the other hand, I like that students don't get penalized or embarrassed for making errors, but rather, like a video game, they get to keep trying until they get it right with no penalty for past failure. I also like that they have at least some control on the pace of their learning in the short term which the current scheme facilitates. I simply became mempals with my students so I could see their overall points and whether they were active users. I used self-reporting for more exact progress monitoring and had them show me their garden only when it was a factor in their grade at midterm. The major complaint and the area that I would suggest for focus is to reduce the time needed to get terms of getting words to the garden. With most words in Esperanto being cognates and without irregularity, one simply does not need as much practice as other languages. In addition, when learning is also taking place in a classroom, through outside reading, etc. rather than just through memrise, using the same algorithm for mastery computation does not make sense. One way to fix this is to have an option for a student to make a word a candidate for harvest when they think they are ready. After a 4 hour wait, if they can type in the word correctly, it could go in the garden (subject to additional watering). If they miss the word, it would go back to sprouting just as if it were not a candidate for the garden. Being able to bypass the sprouting phase for selected words would probably half the time needed to effectively use the system for cognate language that is being studied in a class type setting and make the product more pleasant while still being effective. Overall, Memrise.com, even in its beta state, is an excellent tool for vocabulary acquisition and I recommend it highly.