This simulation study evaluated parameter recovery in cross-classified multiple membership growth curve modeling under a variety of manipulated conditions, including prior distributions (inverse Wishart and uniform distributions), number of measurement occasions (3, 5), number of groups (30, 50, 100), average group size (20, 40), and non-pure hierarchical data rates (20%, 40%). Three Bayesian point estimates were compared: the posterior mean, the posterior median, and the posterior mode. With the inverse Wishart prior, the posterior mean estimates for the between-school variances in the intercept and slope were substantially overestimated with 50 groups or less, while the posterior mode estimates were substantially underestimated. The posterior median estimates were substantially biased with 30 groups or less. With the uniform prior, the posterior mean, posterior median, and posterior mode estimates of the between-school variances in the intercept and slope were substantially overestimated. In general, with the inverse Wishart prior, the Markov chain Monte Carlo (MCMC) posterior median is recommended for cross-classified multiple membership growth curve modeling, except under certain conditions. With the uniform prior, the MCMC mode is recommended for estimating cross-classified multiple membership growth curve modeling rather than the MCMC posterior mean or posterior median.