In order to determine the significance of trisomy mosaicism of an autosome other than chromosomes 13, 18, 20, and 21, 151 such cases diagnosed prenatally through amniocentesis were reviewed. These rare trisomy mosaicism cases include 54 from 17 cytogenetic laboratories, 34 from a previous North American mosaicism survey, and 63 from published reports. All were cases of true mosaicism with information available on pregnancy outcome, and with no evidence of biased ascertainment. There were 11 cases of 46/47, +2; 2 of 46/47, +3; 2 of 46/47, +4; 5 of 46/47, +5; 3 of 46/47, +6; 8 of 46/47, +7; 14 of 46/47, +8; 25 of 46/47, +9; 2 of 46/47, +11; 23 of 46/47, +12; 5 of 46/47, +14; 11 of 46/47, +15; 21 of 46/47, +16; 7 of 46/47, +17; 1 of 46/47, +19; and 11 of 46/47, +22. As to the risk of an abnormal outcome, the data showed a very high risk (> 60 per cent) for 46/47, +2, 46/47, +16, and 46/47, +22; a high risk (40-59 per cent) for 46/47, +5, 46/47, +9, 46/47, +14, and 46/47, +15; a moderately high risk (20-39 per cent) for 46/47, +12; a moderate risk (up to 19 per cent) for 46/47, +7 and 46/47, +7 and 46/47, +8; a low risk for 46/47, +17; and an undetermined risk, due to lack of cases, for the remaining autosomal trisomy mosaics. Most cases were evaluated at birth or at termination, so subtle abnormalities may have escaped detection and developmental retardation was not evaluated at all. Comparison of the phenotypes of prenatally diagnosed abnormal cases and postnatally diagnosed cases with the same diagnosis showed considerable concordance. Since the majority of anomalies noted are prenatally detectable with ultrasound, an ultrasound examination should be performed in all prenatally diagnosed cases. In cytogenetic confirmation studies, the data showed much higher confirmation rates in cases with abnormal outcomes than in cases with normal outcomes [81 per cent vs. 55 per cent for fibroblasts (from skin, fetal tissue, and/or cord); 88 per cent vs. 46 per cent for placental cells; 22 per cent vs. 10 per cent for blood cells]. The confirmation rate reached 85 per cent when both fibroblasts and placental tissues were studied in the same case (with trisomic cells found in one or the other, or both). Therefore, one must emphasize that both fibroblasts and placental tissues should be studied. Except for 46/47, +8 and 46/47, +9, PUBS is of limited value for prenatal diagnosis of rate trisomy mosaicism. DNA studies for UPD are suggested for certain chromosomes with established imprinting effects, such as chromosomes 7, 11, 14, and 15, and perhaps for chromosomes 2 and 16, where imprinting effects are likely.