A method of adjusting for bias due to nonresponse to an alumni survey is demonstrated, based upon analysis of data from both on-time and late respondents. The responses are opinions of 254 alumni regarding how they rate their recently completed program of study within a college of an urban university. As part of the university's curriculum review process, a questionnaire was developed to obtain alumni ratings of their degree program. Twenty-two questionnaire items were selected from the University of Illinois' Program Evaluation Survey (Smock and Hake, 1977), and five demographic questions and a final "would you recommend the program?" question were added. Subjects had received an undergraduate or graduate degree during the last 5 years. It was found that on-time and late respondents differed significantly in their attitudes toward specific aspects of the academic program. Adjustments were made in the average ratings to correct for these differences. In addition, after allowing for random sampling errors, the unadjusted results were contrasted with the adjusted ones. The findings demonstrate that for certain responses, the adjustments were relatively minor, while for other questions the changes were quite significant. In eight cases, the adjusted ratings were significantly less critical of the program than were the unadjusted ratings, based only on the responses of on-time respondents. It is suggested that the increasing reliance on alumni responses to mail questionnaires and their use in the internal review and decision-making process of universities make proper analysis of the data crucial to institutional researchers. The literature on survey nonresponse is also reviewed. (SW)