Both mature and immature mangoes, Mangifera indica L., grown in subtropical Florida are marketed. Food use of immature mangoes includes being cooked for sauce, powdered for a flavoring agent, pickled, and processed into chutney. Immature mangoes are distinguished from mature mangoes in that they do not ripen after harvest (Reid 1992). Immature mangoes have a whitish flesh, lack juiciness, and have a sour flavor, while mature mangoes have yellow flesh, are juicy, and have a sweet flavor. Mango has been listed as a host of Caribbean fruit fly,Anastrepha suspensa (Loew) (Norrbom & Kim 1988). California, where the largest immature mango market is, maintains a Caribbean fruit fly quarantine on both mature and immature imported mangoes. Mangoes can only be shipped from Florida to California if they have undergone an acceptable postharvest treatment that disinfests them of fruit flies. To our knowledge, there have been no studies to determine to what extent immature mangoes may represent a lower risk for fruit fly infestation than mature mangoes. Peina & Moyhuddin (1997) reviewed mature mango fruit resistance toAnastrepha obliqua (Macquart) and infestability differences among cultivars were suggested to be caused by differences in toxic chemicals, nutrients, or resin ducts. Hennessey & Schnell (1995) determined that for mature fruit, some genotypes are more resistant to Caribbean fruit fly than others. They also suggested highly resistant germplasm accessions could be employed in breeding efforts and integrated pest management systems for control of Caribbean fruit flies. The present investigations were conducted to determine if some germplasm accessions in the immature state were more resistant to Caribbean fruit fly than others. Fruits used in the experiments were from the USDA National Clonal Germplasm Repository at the Subtropical Horticulture Research Station in Miami, FL, where over 150 Mangifera accessions are maintained. The accessions comprise a valuable source of genetic diversity utilized by breeders and researchers from all over the world. Fruits from 18 M. indica accessions (individual trees) were selected based on availability and inclusion in a previous study of mature mangoes (Table 1). Immature mangoes were available from March 21 to June 23, 1996. Tree cultivation practices included fertilization, pruning, and chemical weed control. No insecticides were applied during testing and for at least four months prior to testing. The methods for rearing flies, determining mango maturity, and bioassaying of fruits were previously described in detail (Hennessey & Schnell 1995). Flies were reared on an agar-based diet (Hennessey 1994) which was used as the control in the present study. Stage of maturity of the mangoes was determined for each accession and harvest date by holding 5-10 fruits in the laboratory and observing flesh color, juiciness, and flavor after 5-8 days. Each accession was bioassayed on three dates: early (March-April), middle (May), and late (June) immature period (Table 1). Slices, including peel, from five fruits per accession were bioassayed for each date. Ten mature Caribbean fruit fly eggs were inoculated on each fruit slice and corresponding agar control portions. The corrected mean percentage adult emergence was the criterion used to evaluate resistance. This value was obtained by dividing the percentage of eggs reaching adulthood from each slice by the mean percentage reaching adulthood from five agar controls of the same date. The experiment was completely randomized. Corrected mean percentage adult emergence data from the three harvest dates combined (n = 15) were analyzed with PROC ANOVA and compared among cultivars with the LSD test (SAS Institute 1992). The corrected mean percentage of emergence of adult flies from mangoes varied from 0.0% for 'Saigon Seedling' M13269 to 35.7% for'Keitt' (Table 1). The most resistant group included 15 cultivars that supported less than 28.9% emergence. Hennessey & Schnell (1995) observed that the corrected mean percentage emergence varied from 26.6% for mature 'Tobago Small Red' to 119.0%/c (a higher emergence rate than the control) for mature 'Sabre'. In that study, the group that contained the most resistant cultivar (P = 0.10) consisted of 'Peach', 'Irwin', 'Tommy Atkins', 'Rumani', 'Turpentine', 'Keitt', '13-l', 'Sandersha', 'Zilate', and 'Tobago Small Red' (Table 1). Mature mangoes served as a better substrate for development than did immature mangoes of the same cultivar. Over all cultivars, immature mangoes supported from 2-59% of the emergence compared with mature fruits of the same cultivars did. This could be interpreted to mean that, within cultivars, most immature man-