Synthetic control arms (SCA) are increasingly used to assess the comparative effectiveness of treatments in e.g., single arm trials. Bias in the average treatment effect (ATE) estimate due to differences in baseline characteristics between the treatment group and the SCA can be reduced using e.g., inverse probability of treatment weighting (IPTW). In the SCA context, adjusting for too many covariates may result in an estimator with high variance and small effective sample size (ESS). All other things equal, the most prognostic covariates should be selected preferentially for adjustment, but in practice exact prognostic importance is uncertain.