This study used the Learning Potential Test to demonstrate the independence of learning potential from the characteristics of the family environment. The subjects were 36 white children age 46 to 75 months from low and middle socioeconomic level families. Demographic information was obtained by a questionnaire and the home environment quality was evaluated with a revised version of the HOME Inventory (HOME-R). The Learning Potential Test was given individually to children and yielded six scores: spontaneous performance (success without assistance), assisted performance (success with help), transfer performance (performance after assistance, the learning potential score), inefficiency score (part of the assisted performance which did not bring the child to succeed in the transfer task), global performance (score with and without assistance), and gain quotient (proportion of global performance obtained by transfer performance). The Raven's Progressive Colored Matrices were given individually to 21 of the children. Results indicated that the spontaneous performance score correlated with the total HOME-R score and the Raven's scores, suggesting that spontaneous performance is comparable to conventional IQ scores. The transfer performance correlated with only one HOME-R subscale, suggesting that learning potential is independent of the family environment. For families receiving lower HOME-R total scores, the transfer performance was higher than the spontaneous performance, while in the families with higher HOME-R total scores, the transfer performance was lower than the spontaneous performance, suggesting that families with lower home environment quality are more likely to benefit from assistance than families with higher home environment quality. (Contains 49 references.) (KDFB)