New complex deferred annuity contracts render most past consumer comparison efforts simplistic and unrealistic. A questionnaire solicited data from thirty different insurers concerning their annuity offerings. Variable annuity prospectuses and fixed annuity policies are analyzed. Performance measurement techniques are then developed and applied to past and current performance. Portfolio beta analysis as well as a risk-free fee-sensitive efficiency ratio are suggested as methods for evaluating variable annuities, while portfolio rate of return and effective annual yield methods are suggested for analyzing fixed annuity performance. The surge in demand for tax sheltered annuities by employees of 501(c) (3) organizations, by the self employed under the Keogh Act and by those eligible for individual Retirement Accounts, has been met by increasing numbers of insurers offering and constantly modifying qualified plans [1]. Many studies have been conducted for various groups interested in assisting their members make intelligent choices among alternative plans.' In general, however, the prospective purchaser of a tax sheltered annuity is faced with two basic problems: (1) accurate and complete data are not available through any competent, unbiased source; and (2) methods for comparing costs and surveys are only in their formative stages, leaving the prospect with many unanswered questions and often some incorrect impressions. Weisenberger provides some limited data, but not nearly enough to make an informed choice [2]. This study has two main purposes: (1) to reexamine methods currently used to compare alternative annuities, especially in light of changing and unusual practices of insurers; and (2) to suggest improvements in methods of comparison by utilizing some accepted as well as some new financial analysis techniques. It is not the purpose to provide a shoppers guide to tax sheltered annuities. It is assumed that the reader already is familiar with one or more of these types of studies. Concern is with techniques for analysis, rather than current plan comparisons. The paper is divided into three primary sections: techniques for analyzing (1) fixed annuity performance, (2) variable annuity performance, and (3) fee schedules. Jerry D. Todd is Associate Professor of Insurance, Graduate School of Business, The University of Texas at Austin. Several studies have remained unpublished; those published include [3], [4], and [6].