Many poly(ferrocenyldimethylsilane) (PFS) block copolymers form fiber-like micelles with a semicrystalline core in selective solvents. Solvent effects on micelle formation are not well understood. This paper compares micelle formation for a sample of PFS-(50)PI-(1000) (the subscripts refer to the number-average degrees of polymerization) in decane with that in tert-butyl acetate (tBA), a more polar solvent. When micelle formation is seeded, by adding block copolymer as a concentrated solution in tetrahydrofuran to solutions of micelle fragments, micelle growth was similar in both solvents. Micelles with a narrow length distribution were formed and the length increased in proportion to the amount of polymer added. In contrast, when micelles were prepared by heating a sample of the block copolymer in decane or tBA to 90 degrees C and allowing the solution to cool, pronounced differences were observed. In decane, micelles with a uniform width (10 nm) and a length on the order of 5 mu m formed after 1 h, and grew to about 10 mu m after 5 days. In tBA, aliquots taken from solution 1 h after cooling appeared to undergo microphase separation only upon solvent evaporation. Ribbon-like structures were observed after 1 and 5 days aging, but these evolved into fiber-like structures with a uniform 10 nm width and lengths greater than 30 mu m after 25 days. These differences observed in the rate of micelle formation likely reflect differences in the nucleation stage of micelle formation. tBA is a better solvent for the PFS block than decane. As a consequence, it appears to take much longer for semicrystalline micelle nuclei to form in tBA. The seeded growth experiments demonstrate that once seed micelles are present, growth occurs similarly in both solvents.