This study examines the effects of two teaching procedures and two testing procedures (“Skip” and “Guess”) on acquisition, retention and generalization of learning. Three typically developing females between the ages of 8 and 11 learned the 24 lower case letters of the Greek alphabet. Half of the letters were taught with the “Skip” procedure and the other half with the “Guess” procedure. The “Skip” procedure produced faster and more efficient learning than the “Guess” procedure. The “Skip” procedure also resulted in better initial retention (4 weeks), but this effect disappeared in subsequent retention tests. The training conditions did not have differential effects on generalization tests across learning channels, except for the Free/Say channel.