Several experimental groups have recently reported self-assembly of fullerene derivatives on metal substrates. These studies have shown that both the size of the substituent moiety and the presence of hydrogen-bonding functional groups affect the observed adlayer symmetries. However, little theoretical work has been carried out to explain these results. Accordingly, we have carried out classical rigid body Monte Carlo simulations of a variety of fullerene derivatives on a rigid Au(111) surface. We consider several fullerene derivatives functionalized with carboxylic acid groups attached to the fullerene using "arms" of phenyl rings of variable length. A pairwise-additive united-atom potential energy function was constructed using data from a variety of sources. The potential reproduces many of the details observed in experimental studies of C-60 on Au(111). For fullerenes containing two hydrogen-bonding groups, ordered rows of molecules can self-assemble if the fullerenes are able to form a close-packed layer. Steric effects may inhibit close-packing of the fullerenes, resulting in "glassy" adlayers. In addition, the rotational barriers must be sufficiently low that they allow orientational ordering of the molecular adlayer. We have identified two distinct adlayer geometries which can be formed from extended one-dimensional hydrogen-bonding rows. When the barriers to rotation are too high, as is the case when the molecules become sterically hindered, orientational ordering is not possible. Simulations with molecules bearing a single carboxylic acid functionality lead to close-packing and hydrogen-bonded orientational dimers. However, the dimers do not self-assemble into a globally ordered adlayer (with the parametrization used here). In addition, we show that molecules without hydrogen-bonding functional groups can self-assemble into herringbone patterns, provided the conditions for close-packing and orientational ordering have been met. Our findings rationalize several of the experimental results in the literature.