Objective: To determine fellowship satisfaction through a survey of maternal-fetal medicine fellows. Methods: We constructed a survey using multiple choice, Likert scale, ordinal, and categorical scale questions. The questions focused on faculty involvement, mentorship, research time and productivity, education, fellowship selection, ultimate goals of fellows, and satisfaction. We sent the survey in two mailings to all maternal-fetal medicine fellows during April and May 1996. Results: One hundred thirty-eight surveys were mailed, and 136 were returned (98.5% return rate). Twenty-seven percent of fellows did not believe they would complete their thesis by the end of their fellowship. No statistically significant relationship was noted between the fellows’ predicted thesis completion and the availability of funding, support for statistical analysis, the presence of animal research facilities, age, number of dependents, or year of fellowship. The presence of a mentor on the maternal-fetal medicine faculty increased the likelihood of thesis completion from 52.3% to 83.5% (P < .001). Similarly, the presence of a faculty advisor increased the likelihood of thesis completion from 58.9% to 83.5% (P = .001). Thirty-two percent of the respondents did not have a mentor on the faculty, and 41% did not have a faculty advisor. Forty percent indicated that they were too involved in clinical pursuits to perform research. This group was significantly more likely to believe that their theses would not be completed (63% versus 80%, P = .029). Overall, 22% of the fellows would not recommend their fellowships. Fellows with a mentor (88.2% versus 55.8%; P < .001) or faculty advisor (87.3% versus 64.9%; P = .002) were more likely than those without to recommend their fellowship. Conclusion: A mentor or faculty advisor plays a significant role in the training of maternal-fetal medicine fellows and is associated with a higher incidence of satisfaction with the fellowship program, thesis completion, and entrance into academic practice.