An overview of the development and evaluation of an early alert procedure for identifying and assisting students experiencing academic difficulty at Irvine Valley College (IVC), in California, is provided in this report. The first section introduces IVC's new personal computer-based Early Alert Retention System (EARS) software, as well as the research conducted to evaluate the effectiveness of the project as it evolved. The next four sections present research summaries, providing a hypothesis, guiding questions, study methodology, detailed results, tables and charts, and conclusions for the following: (1) a pilot project of faculty mentoring for 160 students randomly selected from a pool of 800 new matriculants and collection of mid-semester alerts from faculty; (2) an evaluation of mid-semester early alerts using an experimental design which compared the performance of students who had been issued a letter of concern/support, those to whom letters requesting that they consult an "early advantage" advisor had been sent, and those not contacted; (3) early alert rosters issued to teachers of all first semester students; and (4) the effectiveness of College Student Inventory for predicting academic outcomes. A key finding presented in the second research summary is that full-time students receiving alert letters had the highest end-of-year retention (81.3%), while part-time students in both the advisor and letter groups had greater end-of-year retention than those not contacted. A detailed discussion of approaches to designing and implementing early alert processes conclude the report. An appendix includes a description of the EARS software, and samples of early alert forms, letters, and logs. (MAB)