I guess this site is pretty famous, since it’s already been on Japan Probe, but it recently got a lot better, so I’d like to write about it here. I’ve been struggling with kanji for 17 years, and Read the Kanji is the first study method I’ve ever found that was truly effective and efficient for learning them.


It’s just a quiz site, but it’s done a thousand times better than all the other quiz programs and kanji books out there. It flashes a kanji on the screen, you type in the pronunciation, and it tells you whether you got it right or wrong. The great thing about Read the Kanji, though, is that it remembers the ones you have trouble with, and once you get a certain number wrong, it goes into “failure review” mode, and you do all the ones you got wrong (you have to set it to “Learning mode” for this to happen. Otherwise it’s a straight out kanji test). Once you get out of failure review mode, those difficult kanji keep coming up again and again until you start to get them right. I find the algorithm very clever about showing me the kanji just often enough that I learn them, but not so often that I get sick of seeing them. I’ve been using the site for a couple months, and I’m learning more kanji in a shorter time than I ever have before.
Read the Kanji has thousands of characters, graded according to the Japanese Language Proficiency Test, and has example sentences with English translations that can be turned on of off as you wish.
The other thing I really like about this site is that the guy behind it clearly considers it a work in progress and is constantly changing and improving it.
The quiz is actually kind of addictive, and I sometimes find it hard to quit once I get started, a problem I’ve never had when studying kanji in the past!

Read the Kanji: http://www.readthekanji.com/quiz

Just a note to all of the people who are coming from the Heisig site (http://forum.koohii.com/viewtopic.php?id=2958&action=new). I can read kanji well enough to do a fair amount of professional translation work, and am familiar with the Heisig method, but didn’t find it worked for me.