The effectiveness of developmental education policies at a comprehensive community college was investigated using a new student tracking system. A sample of 1,644 students were tracked from fall 1992 until the end of the fall 1994 semester, 1,226 of whom evidenced basic skills and were eligible for college-level courses, and 418 of whom were classified as skill-deficient. After the fall 1994 semester, 179 skill-deficient students had not remediated and 239 students had remediated. The college-level and skill-deficient students were compared on persistence and performance. Also examined were links between skill-deficient students' academic performance and their enrollment patterns in developmental education courses exclusively, college-level courses and remedial courses, and college-level courses only. Additional concerns were rates of remediation and types of problems (math only, language only, or both, and reading and/or writing deficiency, or both). Results support the following recommendations: skill-deficient students should be required to remediate upon initial enrollment; these students should be allowed to take college-level courses along with remedial studies, but language deficient and triple-deficient students should be encouraged to finish remediation before beginning college-level courses. In conclusion, the results point to guidelines for designing or revising policies governing developmental education programs. (Contains 11 references.) (SW)