A study examined whether children learn phonic decoding skills by reading without direct phonic instruction; compared the effects of a whole language first grade reading program with the effects of a traditional basal reading program; and determined whether there was a difference in decoding and comprehending abilities across levels of ability. Subjects, an experimental group of 33 low socioeconomic first grade students in two classrooms who completed the year-long program and a control group of 33 low socioeconomic students in two other elementary schools matched by sex and reading readiness, completed a Nonsense Word Test to assess decoding ability and a Deletion test to assess their ability to comprehend increasingly more difficult prose. The control group used the Scott Foresman Basal Series. Testing occurred upon completion of the school year. Comparison of the test results indicated that: (1) students in the whole language classrooms scored as well in decoding ability as those students who received direct phonics instruction; (2) as a group, students in the whole language classrooms scored higher on the Deletion Test than the control group (indicating greater comprehension of increasingly difficult prose); although (3) subjects scoring in the middle and lowest on a reading readiness test accounted for much of the difference in the overall difference in the mean scores on the Deletion Test. Findings suggest support for the whole language approach. (RS)