We report observations of nine main-belt comets (MBCs) or candidate MBCs, most of which were obtained when the targets were apparently inactive. We find effective nucleus radii (assuming albedos of p
V
= 0.05 ± 0.02) of r
n
= (0.24 ± 0.05) km for 238P/Read, r
n
= (0.9 ± 0.2) km for 313P/Gibbs, r
n
= (0.6 ± 0.1) km for 324P/La Sagra, r
n
= (1.0 ± 0.2) km for 426P/PANSTARRS, r
n
= (0.5 ± 0.1) km for 427P/ATLAS, r
n
< (0.3 ± 0.1) km for P/2016 J1-A (PANSTARRS), r
n
< (0.17 ± 0.04) km for P/2016 J1-B (PANSTARRS), r
n
≤ (0.5 ± 0.2) km for P/2017 S9 (PANSTARRS), recently redesignated 455P/PANSTARRS, and r
n
= (0.4 ± 0.1) km for P/2019 A3 (PANSTARRS). We identify evidence of activity in observations of 238P in 2021, and find similar inferred activity onset times and net initial mass-loss rates for 238P during perihelion approaches in 2010, 2016, and 2021. P/2016 J1-A and P/2016 J1-B are also found to be active in 2021 and 2022, making them collectively the tenth MBC confirmed to be recurrently active near perihelion and therefore likely to be exhibiting sublimation-driven activity. The nucleus of 313P is found to have colors of
g
′
−
r
′
=
0.52
±
0.05
and
r
′
−
i
′
=
0.22
±
0.07
, consistent with 313P being a Lixiaohua family member. We also report nondetections of P/2015 X6 (PANSTARRS), where we conclude that its current nucleus size is likely below our detection limits (r
n
≲ 0.3 km). Lastly, we find that of 17 MBCs or candidate MBCs for which nucleus sizes (or inferred parent body sizes) have been estimated, >80% have r
n
≤ 1.0 km, pointing to an apparent physical preference toward small MBCs, where we suggest that Yarkovsky–O’Keefe–Radzievskii–Paddack spin-up may play a significant role in triggering and/or facilitating MBC activity.